copyright and use of this thesis · other degree or diploma of the university or other institute of...

Copyright and use of this thesis This thesis must be used in accordance with the provisions of the Copyright Act 1968. Reproduction of material protected by copyright may be an infringement of copyright and copyright owners may be entitled to take legal action against persons who infringe their copyright. Section 51 (2) of the Copyright Act permits an authorized officer of a university library or archives to provide a copy (by communication or otherwise) of an unpublished thesis kept in the library or archives, to a person who satisfies the authorized officer that he or she requires the reproduction for the purposes of research or study. The Copyright Act grants the creator of a work a number of moral rights, specifically the right of attribution, the right against false attribution and the right of integrity. You may infringe the author’s moral rights if you: - fail to acknowledge the author of this thesis if you quote sections from the work - attribute this thesis to another author - subject this thesis to derogatory treatment which may prejudice the author’s reputation For further information contact the University’s Director of Copyright Services sydney.edu.au/copyright

Soil carbon stocks under different forest types in Bhutan, Eastern

Himalayas

By

Sonam Tashi

A thesis submitted in fulfilment of the requirements for the degree of

Doctor of Philosophy

Centre for Carbon, Water and Food

Faculty of Agriculture and Environment

The University of Sydney

2017

i

Certificate of Originality

I hereby declare that this submission is my own work and that, to the best of my

knowledge and belief, it contains no material previously published or written by another

person nor material which to a substantial extent has been accepted for the award of any

other degree or diploma of the university or other institute of higher learning, except where

due acknowledgment has been made in the text.

Signature: Sonam Tashi

Date: 31/08/2016

ii

Statement of Authorship Tashi, S., Singh, B., Keitel, C., Adams, M., 2016. Soil carbon and nitrogen stocks

in forests along an altitudinal gradient in the eastern Himalayas and a meta-analysis of

global data. Global Change Biology 22(6), 2255-2268 (Chapter 3).

Tashi, S., Keitel, C., Singh, B., Adams, M., Elevation and light drive abundances

of carbon and nitrogen isotopes in soil and vegetation in the Himalayas. Ecosystems −

Submitted (Chapter 4).

Tashi, S., Keitel, C., Singh, B., Adams, M., Allometric equations for biomass and

carbon stocks of forests along an altitudinal gradient in the eastern Himalayas. Forestry −

Accepted (Chapter 6).

Sonam Tashi was responsible for carrying out the field work, completing the

laboratory and data analysis, drafting and submitting the manuscript to the journal.

Professor Balwant Singh, supervised the entire process of the study and

contributed with editorial advice

Dr. Claudia Keitel, supervised the entire process of the study, helped with

laboratory analysis and provided editorial advice

Professor Mark Adams, contribute to discussions and provided helpful comments

and editorial advice.

I hereby certify that the above statement about my contribution to the research

work in this PhD thesis is true and accurate, and I give Sonam Tashi full permission to

submit this work as part of his PhD thesis.

Balwant Singh Claudia Keitel Mark Adams

iii

Acknowledgments My PhD program was possible due to an accumulation of efforts by many

individuals and organizations. I would like to express my sincere gratitude to all of them

and will always remain indebted. Foremost, I would like to express my special

appreciation and gratitude towards my supervisor Professor Balwant Singh, who made

himself generously available to shed his insights and immense knowledge in guiding my

research. He has always been very supportive of my entire research project; sourcing funds

for my field works, encouraging me to attend conferences and providing valuable input

during my thesis write up.

I would also like to thank Dr. Claudia Keitel for her exceptional supervision with

regards to isotopic analysis. I am grateful to her for introducing me to the possibilities and

potential of isotopic signals in our environment. I also take the opportunity to thank

Professor Mark Adams for his support, guidance and exploring funds for my field works in

Bhutan. I am extremely grateful to all my supervisors, without whom I am sure that this

project of mine would have remained futile.

I take the opportunity to thank the academics and technical staffs within the

University for their support. In particular, I would like to thank Floris van Ogtrop for

providing comprehensive support in designing my field work and data analysis; Lori

Watson, Michael Turner and Janani Vimalathithen for their technical support and guidance

with my laboratory analysis. At the University, I would like to thank my postgraduate

colleagues both past and current for making my stay here in Sydney a most memorial one.

In particular I would like to thank Dr. Tshering Dorji, Dr. Jonathan Mangmang, Dr.

Nirmala Liyanage, Claudia Carrasco Cabrera, Niranjan Manikku and Dr. Ali Khoddami

for making our lunch sessions lively with lots of discussions. I would also like to thank

my mates Sabina Yeasmin and Alexandra Keith for their friendship and providing

comments and suggestions for my project. Further I would like to thank Kanika Singh,

Patrick Filippi, Edward Jones, William Salter and Alexandra Barlow for their

companionship.

iv

To the Bhutanese community here in Sydney, I would like to convey my deepest

gratitude for those memorial gatherings organized. Specifically, I would like to

acknowledge Dr. Sherub Phuntsho, Aum Tashi Lhamo, Dr. Tshering Dorji, Kiba Choden

and family, Dr. Ratna Bdr. Gurung and family, Dr.Tenzin and family, Gyembo Sithey,

Pem Dem and family, Rinchen, Sonam Yangdon and family, Ugyen Lhendup, Kezang

Dema and family for the great support and company.

I would also like to take this opportunity to thank the Ministry of Agriculture and

Forests for their support during the entire duration of my project. In particular, I would like

to thank Dr. Kinley Tenzin and Dr. Purna Bdr.Chhetri for providing logistic support for

the entire duration of the field work. I am indebted to the superb team of field crew, Dawa

Tshering, Sonam Tobgay, Yograj Chhetri, Kunzang Dhendup, Tanka, Harka, Tshering

Wangchuk and Ugyen. Without their support, talent and dedication, I would never have

been able to complete all my field work.

I would like to expresses my gratitude for the generous funding provided by the

Australian Government through the Endeavour Postgraduate Scholarship for my PhD

program. I thank Ms Shireen Ravesteyn, my case manager with Scope Global Pty Ltd.,

who efficiently execution all the affairs related to my scholarship.

To my family I am eternally grateful. I thank my brothers Colonel Lhatu

Tshering and family, Major Tshering Dorji and family and my sisters Kelzang Lhadon and

Nima for their love and support for every endeavour I undertake. I would also like to thank

my father in-law Brigadier Dal Bhadur Chhetri (retd) and my late mother in law Mrs

Bishnu Maya Chhetri for all the love. Finally I would like to thank my Mum Dechen

Wangmo, who has always been the source of my inspiration, my wife Mumta Chhetri,

who has always been the source of my strength and my two daughters Dechen Yangzom

and Kinzang Sonam Tshomo, who are very reasons that I take up such challenges.

Tashi Delek

v

Publications

Tashi, S., Singh, B., Keitel, C., Adams, M., 2016. Soil carbon and nitrogen stocks in forests along an altitudinal gradient in the eastern Himalayas and a meta-analysis of global data. Global Change Biology 22(6), 2255-2268. doi.10.1111/gcb.13234 (Chapter 3).

Tashi, S., Keitel, C., Singh, B., Adams, M., Elevation and light drive abundances of carbon and nitrogen isotopes in soil and vegetation in the Himalayas. Ecosystems − Submitted (Chapter 4)

Tashi, S., Keitel, C., Singh, B., Adams, M., Allometric equations for biomass and carbon stocks of forests along an altitudinal gradient in the eastern Himalayas. Forestry − Accepted (Chapter 6).

Conference oral presentation:

Tashi, S., Keitel, C., Singh, B., Adams, M. The effect of altitude on C and N dynamics in biomass and soil in the eastern Himalaya. 13th Australasian Environmental Isotope Conference, Sydney 8 – 10 July, 2015.

Tashi, S., Vulnerable Mountain forest soils. Three minute thesis competition. Faculty of Agriculture and Environment Research Symposium, The University of Sydney, New South Wales, Australia July 12, 2016

Conference Poster Presentations:

Tashi, S., Singh, B., Keitel, C., Adams, M., The assessment of total carbon and nitrogen stocks along an altitudinal gradient in the eastern Himalayas. Faculty of Agriculture and Environment Research Symposium, The University of Sydney, New South Wales, Australia July 14, 2015

Tashi, S., Singh, B., Keitel, C., Adams, M., Soil carbon stock along the foothill of the Himalayas. National Soil Science Conference, MCG, Melbourne, Victoria 23 – 27 November, 2014.

vi

I dedicate this thesis to my Mum

Dechen Wangmo

vii

Abstract

Climate change is one of the greatest current challenges for humankind. We have

pushed carbon dioxide (CO2) concentration in the atmosphere from preindustrial 280 ppm

to 404.3 ppm in July 2016. With the increase in greenhouse gases in the atmosphere, we

experience more frequent and severe weather with devastating consequences to human

lives and the environment. To tackle this issue, the United Nations Framework Convention

on Climate Change (UNFCCC) initiated mitigation strategies to reduce emissions of CO2,

which is one of the six greenhouse gases listed in the Kyoto Protocol. As a member

country to the UNFCCC, Bhutan is obliged to monitor greenhouse gas emissions and

report possible sources and sinks. There is a lack of comprehensive research on carbon (C)

sinks and emissions in Bhutan, and this dissertation focusses on quantifying C stocks from

the Eastern Himalayan forest ecosystems (biomass and soils) along an altitudinal gradient

from 317 to 3300 m.

Firstly, this research investigates C and nitrogen (N) dynamics in soils under high

altitude forests which potentially store a large pool of C and N. Nitrogen is an essential

mineral nutrient for plants and therefore closely related to the C dynamics. Total soil C and

N stocks significantly increased with altitude and decreased with soil depth. Carbon and N

stocks were significantly correlated with altitude (as a proxy for environmental conditions)

which accounted for 73% and 47% of the variation in C and N stocks, respectively.

Temperature and altitude had similar correlation coefficients and temperature was

ostensibly the main driver of soil C and N along the altitudinal gradient. Increasing soil C

and N stocks were associated with forest composition, forest basal area (BA) and quantity

of leaf litter which in turn was influenced by altitude and temperature.

To elucidate the driving processes of C and N stocks, inputs, turnover and

stability, C and N isotopes in soil and biomass were measured. It was established that

overstorey vegetation contributes significantly to the soil C, as δ13C of overstorey and soil

showed similar trends along the altitudinal gradient. δ13C and δ15N enrichment with soil

viii

depth was least for highest altitude forest. Additionally, the slope of soil δ13C versus the C

concentration, which is indicative of organic matter decomposition, was also smallest at

the highest altitude forest. This suggests slow turnover of C and N in the high altitude

forest soils, which was further supported by increased C:N ratio and CEC with increasing

altitude.

The decomposition of OM proceeds via complex biological, physical and

chemical processes and resulting in associations with minerals in the soil. Sequential

density fractionation, DRIFT spectroscopy and IRMS were used to determine the different

proportion and forms of C in forest soils. Lighter soil density fractions had a greater

proportion of aliphatic C that were largely associated with phyllosilicates, while the

heavier soil density fractions had a greater proportion of aromatic C that was usually

associated with quartz. The larger proportion of aromatic C in the higher soil density

fractions suggests that SOC in this fraction has been more processed, corroborated by the

accompanied decreased C:N ratio and enrichment of δ13C with increasing soil density

fractions.

This study depicts reduced decomposition in soils at higher altitude forests with

proportionally greater aliphatic C to aromatic C, than at lower altitudes, coupled with

increasing C:N ratios for all soil density fractions with increasing altitude. In addition, the

ratio of C to O functional groups, which is a measure of relative recalcitrance of organic

carbon (OC), for the highest altitude is low, signifying limited decomposition even for

easily decomposable carboxyl and polysaccharides. If global warming continues unabated,

the large C stocks in mountainous regions that are predominantly in labile form could be

an additional source of CO2 and further aggravate global warming.

Soil organic carbon may be the largest terrestrial pool of C, but it is closely

associated with land use and land cover. In a forest eco-system, the aboveground biomass

is not only an important source of SOC, but also stores large amounts of C in the trunks

and branches. Aboveground biomass (AGB) allometric equations were developed to

estimate forest AGB C stocks for the study area as there are limited equations available for

ix

the Himalayas. To construct allometric equations, 144 trees were harvested with diameters

ranging from 10 cm to 77 cm, from the five different forest types found along the

altitudinal study transect. Model selection was based on the Akaike Information Criterion

(AIC), root mean square error (RMSE), coefficient of determination (r2) of the regression

and absolute average deviation from the measured AGB. Tree diameter at breast height

(DBH), height and wood specific gravity (WSG) were the variables used to build the

models. Two forms of models were identified that could predict AGB across a range of

trees using DBH and tree height. Although the inclusion of WSG in the model improved

the AGB prediction, it is recommended to use models that consider DBH and height of

trees. Wood specific gravity is not collected during conventional forest inventory and data

may not be available for all the tree species. For the five forest types in the study area,

specific allometric biomass equations were developed. Using the best-fit models, estimated

AGB C stocks increased with altitude from 57 to 207 Mg C ha-1. The use of measured C

concentration rather than an assumed 50% C for biomass reduced estimated AGB C stocks

between 6.8 and 8.6%. With this study, baseline data for soil C and N stocks and

allometric equations for biomass C stock estimation were developed. The estimation of C

stocks in the forest soils and biomass allometric equations for the different forest types in

the Bhutan Himalayas will enable the region to better monitor its C stocks and emission to

benefit from the United Nations REDD programs.

x

Contents Certificate of Originality ...................................................................................................... i

Statement of Authorship ..................................................................................................... ii

Acknowledgments ............................................................................................................... iii

Publications .......................................................................................................................... v

Abstract .............................................................................................................................. vii

Contents ................................................................................................................................ x

List of Tables ..................................................................................................................... xvi

List of Figures .................................................................................................................. xvii

List of Supplementary ...................................................................................................... xix

Chapter 1. Introduction ...................................................................................................... 1

1.1 Aims ...................................................................................................................... 3

1.2 Thesis outline ........................................................................................................ 4

Chapter 2. Literature Review ............................................................................................. 9

2.1 Forest carbon dynamics ......................................................................................... 9

2.1.1 Soil organic carbon in terrestrial ecosystems ........................................ 11

2.1.2 Preservation of soil organic carbon ....................................................... 13

2.1.3 Characterisation of soil organic carbon ................................................ 14

2.1.4 Factors influencing SOC stocks ............................................................ 15

2.1.5 Forest biomass carbon ........................................................................... 16

2.1.6 Measurement of the carbon pool in above ground tree biomass .......... 17

2.1.7 Carbon pool in below ground root biomass .......................................... 19

2.1.8 Carbon pool in coarse woody debris ..................................................... 21

xi

2.1.9 Carbon pool in coarse leaf litter ............................................................ 22

2.1.10 Carbon content in biomass .................................................................... 23

2.2 Summary ............................................................................................................. 23

Chapter 3. Soil carbon and nitrogen stocks in forests along an altitudinal gradient in

the eastern Himalayas and a meta-analysis of global data ............................................ 41

Abstract ......................................................................................................................... 41

3.1 Introduction ......................................................................................................... 42

3.2 Material and Methods ......................................................................................... 44

3.2.1 Study area .............................................................................................. 44

3.2.2 Soil sampling and analysis .................................................................... 46

3.2.3 Forest inventory, biomass sampling and analysis ................................. 47

3.2.4 Statistical analysis ................................................................................. 48

3.2.5 Meta-analysis for soil C along the altitudinal gradient ......................... 48

3.3 Results ................................................................................................................. 49

3.3.1 Influence of forest type and soil depth on carbon and nitrogen stocks in

the soil on the Bhutan transect. ............................................................. 49

3.3.2 Influence of forest type and altitude on carbon and nitrogen

concentrations in understorey live biomass, canopy dead wood and leaf

litter ................................................................................................... 53

3.3.3 Influence of edaphic parameters on the carbon and nitrogen stocks in

soil ................................................................................................... 56

3.3.4 Meta-analysis of studies on soil carbon trend along altitudinal gradients.58

3.4 Discussion ........................................................................................................... 60

3.4.1 Soil carbon and nitrogen trends with altitude ....................................... 60

3.4.2 Biomass carbon concentration with altitude ......................................... 62

xii

Chapter 4. Elevation and light drive abundances of carbon and nitrogen isotopes in

soil and vegetation in the Himalayas. ............................................................................... 75

Abstract ......................................................................................................................... 75

4.1 Introduction ......................................................................................................... 76

4.2 Material and Methods ......................................................................................... 77

4.2.1 Site description ...................................................................................... 77

4.2.2 Plant sampling ....................................................................................... 79

4.2.3 Soil sampling ........................................................................................ 79

4.2.4 Stomatal measurements and calculation of gw max ................................. 80

4.2.5 Isotope and elemental analysis .............................................................. 80

4.2.6 Carbon and nitrogen isotopic enrichment with depth ........................... 81

4.2.7 Statistical analysis ................................................................................. 81

4.3 Results ................................................................................................................. 81

4.3.1 Biomass carbon and nitrogen isotope trends with forest types ............. 81

4.3.2 Biomass carbon and nitrogen isotope trends with altitude ................... 83

4.3.3 Tree stomatal density and conductance with altitude and forest types . 86

4.3.4 Soil carbon and nitrogen isotope trends with forest type, altitude and

soil depth. .............................................................................................. 87

4.3.5 C:N ratio in the soil along the altitudinal gradient ................................ 88

4.3.6 Relationship of carbon and nitrogen isotopes to total C and N

concentrations in soil ............................................................................ 89

4.3.7 Correlations between soil and biomass C and N isotopes and soil

properties ............................................................................................... 91

4.4 Discussion ........................................................................................................... 93

xiii

4.4.1 Biomass and soil carbon isotope trends with forest type and altitude .. 93

4.4.2 Biomass and soil nitrogen isotopes trends along the altitudinal gradient.95

Chapter 5. Mineral-organic associations and organic carbon forms in forest soils at

different altitudes of eastern Himalayas ........................................................................ 110

Abstract ....................................................................................................................... 110

5.1.1 Introduction ......................................................................................... 111

5.2 Materials and Methods: ..................................................................................... 112

5.2.1 Study area ............................................................................................ 112

5.2.2 Soil density fractionation .................................................................... 114

5.2.3 Isotopic analysis .................................................................................. 114

5.2.4 Mineralogical analysis of soils ............................................................ 115

5.2.5 Spectroscopic analysis of soil density fractions .................................. 115

5.2.6 Statistical Analysis .............................................................................. 116

5.3 Results ............................................................................................................... 117

5.3.1 Characterization of the different altitude forest soils .......................... 117

5.3.2 Properties for soil density fractions .................................................... 119

5.3.3 DRIFT analysis ................................................................................... 123

5.4 Discussion ......................................................................................................... 130

5.4.1 Relationship between soil properties, soil density fractions and various

altitude soils ........................................................................................ 130

5.4.2 Organo-mineral association ................................................................ 131

xiv

Chapter 6.Allometric equation for biomass and carbon stocks of forest along an

altitudinal gradient in the eastern Himalayas ............................................................... 141

Abstract ....................................................................................................................... 141

6.1 Introduction ....................................................................................................... 142

6.2 Material and methods ........................................................................................ 143

6.2.1 Study site ............................................................................................. 143

6.2.2 Forest inventory and zonation ............................................................. 143

6.2.3 Tree biomass data collection ............................................................... 145

6.2.4 Measurement of specific gravity of wood cores ................................. 146

6.2.5 Calculation of biomass in the understorey vegetation ........................ 146

6.2.6 Models and statistical analysis ............................................................ 147

6.2.7 Comparison of model selected for each forest type with previously

published equations ............................................................................. 149

6.3 Results ............................................................................................................... 150

6.3.1 Carbon concentrations in overstorey tree wood, leaves and specific

gravity of wood ................................................................................... 150

6.3.2 Model selection ................................................................................... 151

6.3.3 Model selection for the tropical forest ................................................ 151

6.3.4 Model selection for the sub-tropical forest ......................................... 151

6.3.5 Model selection for the warm tropical broadleaved forest ................. 152

6.3.6 Model selection for the cool temperate broadleaved forest ................ 152

6.3.7 Model selection for the cold temperate forest ..................................... 152

6.3.8 Model selection for the entire forest ................................................... 153

6.3.9 Model comparison to published equations.......................................... 158

xv

6.3.10 Aboveground overstorey and understorey biomass of the different

forest types .......................................................................................... 159

6.4 Discussion ......................................................................................................... 160

6.4.1 Biomass carbon concentration and aboveground biomass in the

different forest types ........................................................................... 160

6.4.2 Aboveground tree biomass model selection ....................................... 161

6.4.3 Comparison of models to various other models ................................. 162

Chapter 7. General Discussion and Conclusions .......................................................... 178

xvi

List of Tables Table 3.1 Forest characteristics along the altitudinal gradient ............................................ 45

Table 3.2 Total carbon and nitrogen stocks (Mg ha-1) and percentages of total at various

depths in soils .............................................................................................................. 51

Table 3.3 Nitrogen and carbon contents on mass (%) and volume ..................................... 54

Table 3.4. Pearson correlation coefficients between soil and biomass properties .............. 59

Table 4.1 Climate and forest characteristics along the altitudinal gradient ........................ 78

Table 4.2 Carbon and nitrogen isotopes in different biomass categories for different forest

zones. ........................................................................................................................... 82

Table 4.3 Pearson correlation coefficients between carbon and nitrogen isotopes and

concentrations in soil ................................................................................................... 92

Table 5.1 Physico-chemical properties of soils from the two top genetic horizons of soil

profiles from forests at different altitudes in Bhutan. ................................................ 113

Table 5.2 DRIFT spectra band assignment for organic and inorganic bands.. ................. 116

Table 5.3 Pearson correlations for DRIFT bands representing organic and inorganic

functional groups and soil properties ......................................................................... 129

Table 6.1 Characteristics of forest along the altitudinal gradient. .................................... 144

Table 6.2 Carbon content on mass (%) basis and specific gravity of wood in the overstorey

biomass from different forest zones in Bhutan .......................................................... 150

Table 6.3 List of models developed for estimation of aboveground tree biomass for the

different forest types in Bhutan. ................................................................................ 154

Table 6.4 Mean RMSE, r2 and % bias for training and validation data set for the different

models used for various forest types. ........................................................................ 155

Table 6.5 Confidence interval of the measured mean AGB and average deviation of

estimated AGB using models of best fit developed for the different forest types. .... 158

Table 6.6 Total measured aboveground biomass and carbon from various biomass

components for different forest types. ....................................................................... 160

xvii

List of Figures

Fig. 2.1 Total carbon stocks in different terrestrial components. ........................................ 13

Fig. 3.1 Geographic location of the study area with main land use land cover types ......... 45

Fig. 3.2 Relationship between total C and N stocks with altitude for soil depths from 0 –

30 cm (a and b) and 0 – 100 cm (c and d). .................................................................. 50

Fig. 3.3 Relationship between soil C:N ratio and altitude for different soil depth categories

..................................................................................................................................... 52

Fig. 3.4 Average carbon stocks in the present study and from other forest ecosystems

around the world. ......................................................................................................... 52

Fig. 3.5 Carbon concentration in (a) dead wood, (b) leaf litter, (c) understory foliage and

(d) stem wood at different altitudes.. ........................................................................... 55

Fig. 3.6 Relationship between tree biomass with altitude ................................................... 56

Fig. 3.7 Relationship between C and N stock with leaf litter, clay percent, CEC, soil C:N

ratio and bulk density.. ................................................................................................. 57

Fig. 3.8 (a) Effect size (Zr) for Fisher’s z and 95% CI for SOC with altitude for individual

studies. ......................................................................................................................... 58

Fig. 4.1 Differences of δ13C in overstorey leaves to other biomass components. .............. 83

Fig. 4.2 Altitudinal trend of δ13C in overstorey biomass..................................................... 84

Fig. 4.3 Altitudinal trend of δ15N in overstorey biomass .................................................... 85

Fig. 4.4 Relationship of maximum potential stomatal conductance with δ13C in sunlit

overstorey leaves. ........................................................................................................ 87

Fig. 4.5 δ13C ± S.D (a) and δ 15N ± S.D (b) at different soil depths for forest types in the

study area. .................................................................................................................... 88

Fig. 4.6 Soil carbon to nitrogen ratio at different depths and under different forest types

along the altitudinal gradient. ...................................................................................... 89

xviii

Fig. 4.7 Relationship between δ13C and log transformed C concentration (g kg-1, a − e),

and δ15N and log transformed N concentration (g kg-1,f − j) in soils with depth ........ 90

Fig. 5.1 Random powder X ray diffraction patterns of different altitude forest soils a)

surface soils b) sub-surface soils. .............................................................................. 117

Fig. 5.2 X-ray diffraction patterns of the oriented clay fractions of different altitude forest

soils ........................................................................................................................... 118

Fig. 5.3 Proportion of soil mass, total C and total N distribution in density fractions ...... 120

Fig. 5.4 Total soil (a & b) C and (c & d) N concentrations, (e & f) C:N ratios and (g & h)

δ13C values in different density fractions for various altitude forest soils. ................ 122

Fig. 5.5 DRIFT spectra of the four density fractions of surface soils ............................... 124

Fig. 5.6 Relative integrated peak area of organic bands in the DRIFT spectra density

fractions ..................................................................................................................... 126

Fig. 5.7 Indices calculated from the relative integrated peak area of organic bands in the

DRIFT spectra of density fractions. ........................................................................... 127

Fig. 6.1 The relationship between observed and predicted total above ground tree biomass

with tree DBH ............................................................................................................ 157

xix

List of Supplementary Supplement 3.1 Data sources used in the meta-analysis that provided correlation data on

SOC with altitude, MAT, MAP, BA and forest floor leaf litter. ................................. 72

Supplement 3.2 Soil properties at different soil depths for forest types in Bhutan. ........... 73

Supplement 3.3 Total carbon and nitrogen stocks in Mg ha-1 at various depths in soils

from different forest zones in Bhutan. ......................................................................... 74

Supplement 4.1 Leaf stomatal measurements for Persea sp.. ......................................... 107

Supplement 4.2 Leaf stomatal measurements for tree species ....................................... 107

Supplement 4.3 Relationship of maximum potential stomatal conductance Gw max with

altitude ....................................................................................................................... 107

Supplement 4.4 Relationship of the difference of δ15N in the understorey leaf and soil

with altitude. .............................................................................................................. 108

Supplement 5.1 Random powder X ray diffraction patterns of different density fractions

of soils from ............................................................................................................... 140

Supplement 6.1 Scatter plots for measured log tree biomass versus log DBH .............. 170

Supplement 6.2 Tree species specific gravity of wood and C content ............................ 170

Supplement 6.3 Discrepancies in aboveground biomass C stock estimation .................. 171

Supplement 6.4 Biomass weight of individual tree components of the different forest

types. .......................................................................................................................... 172

Chapter 1. General Introduction

Page | 1

Chapter 1.

Introduction

The global concentration of carbon dioxide (CO2) in the atmosphere has

risen steadily from 280 parts per million (ppm) in 1750 to 404.3 ppm in July 2016

(www.esrl.noaa.gov/gmd/ccgg/trends). Increases in greenhouse gas (GHG)

emissions are considered to be the main driver of global warming. The burning of

fossil fuels (Le Quere et al., 2009), deforestation and conversion of land for

agricultural use (Murty et al., 2002) are major anthropogenic sources of atmospheric

CO2. Globally, the area of tropical forests for the year 2010 was 1541 million ha and

this reduced by 7.4% compared to the estimated area in 1990 (Achard et al., 2014).

Tropical deforestation alone is estimated to cause one-fourth of the anthropogenic

carbon (C) emissions (Kindermann et al., 2008). International organizations, such as

the United Nations Framework Convention for Climate Change (UNFCCC) are

therefore considering actions for reducing emissions from deforestation and forest

degradation (REDD) programs. The Intergovernmental Panel on Climate Change

(IPCC) estimates between 0.9 and 4.3 Gt of C are absorbed annually in soil and

vegetation (Stocker et al., 2013), and as forests play a vital role in increasing the C

sink and reducing C emissions, it is imperative to assess and monitor terrestrial C

stores in forests. The premise of a future REDD policy is financial compensation for

countries willing and capable to reduce emissions by mitigating deforestation and

forest degradation. However, in order to benefit from REDD programs, countries

must have reliable baseline data and be capable of monitoring the rate of change of

CO2 emissions. The accessibility to financial gains by REDD programs may initiate a

paradigm shift in forest management for both economic and environmental benefits.

Carbon stocks in the soils are dependent upon numerous factors, like land

use (Dorji et al., 2014), soil type (Bauhus et al., 1998), forest type (Deng et al.,

2009), climatic conditions (Jobbagy and Jackson, 2000) and topography (Lal, 2005;


Page | 2

Prichard et al., 2000). Although globally, numerous studies have been carried out to

estimate soil C stocks, extrapolation of C data from other regions is difficult, as no

two sites are identical. Zhang et al. (2011) reported that in Changbai Mountains of

China there is no obvious altitudinal pattern in SOC distribution, while Martin et al.

(2010) found that altitude and climate predominantly influence C storage in the soil

at higher altitudes (>1700 m.a.s.l.), whereas at lower altitudes (900-1700 m.a.s.l)

vegetation and land form were the dominant influence. Although numerous studies

for SOC distribution exist in various ecosystems, equally numerous conclusions have

been drawn. Not many studies have focused on the content and distribution of SOC

fractions along different altitudes in forest ecosystems (Zhang et al., 2011) and no

such data are available for different forests in Bhutan Himalayas.

Forest biomass represents another important C pool. Estimation of the

current and future potential of C sequestration in forests is dependent upon temporal

change and on the age and species of the forest stand (Mendoza-Ponce and Galicia,

2010; Somogyi et al., 2007). The current method of estimating biomass stock uses

inventory data multiplied by a carbon fraction to establish the corresponding carbon

stock (Somogyi et al., 2007), although various alternative methods are emerging,

including aerial photogrammetry and the emerging use of light detection and ranging

(LiDAR).

It has been recognised that the highest accuracy in estimating the forest

biomass is obtained using a biomass equation that represents individual tree species

(Petersson et al., 2012). This suggests that stand and species specific biomass models

must be developed to reduce estimation errors.

Comprehensive databases for species specific and stand biomass and

volume equations are available for the pan tropics (Chave et al., 2005), Europe

(Muukkonen, 2007; Muukkonen et al., 2005; Zianis et al., 2005), Australia (Eamus et

al., 2000; Kieth et al., 1999; Snowdon et al., 2000) and the North Americas

(Chojnacky et al., 2014; Ter-Mikaelian and Korzukhin, 1997). For the Himalayan


Page | 3

region, a few allometric equations have been developed for species in the western

(Garkoti, 2008) and central (Negi et al., 1983; Rana et al., 1989) areas based on

limited sample size and some for small diameter trees only (Singh et al., 2011). As

these biomass equations are stand- and site-specific, they are unreliable to use for

other regions (Jenkins et al., 2003).

1.1 Aims

There is a clear lack of reliable C stock data in soils and vegetation in the

Himalayan region. Therefore this study aimed to develop and estimate reliable

standards for forest C stocks in the Bhutan Himalayas. The study area encompassed

five major forest types found along an altitudinal gradient in the foothills of the

Eastern Himalayan belt.

The specific aims of the project were to:

i. quantify the total soil C and N stocks at different depths along an

altitudinal gradient in Bhutan and to synthesize a global relationship

between soil C and altitude from different eco-regions of the world

via a meta-analysis.

ii. identify factors that determine patterns in δ13C and δ15N in biomass

and soil with altitude and identify the drivers of C and N input,

turnover and stability in the soil.

iii. characterize C forms associated with different minerals in the SOM

and determine the proportion of C forms associated with different

minerals for the different altitude forest soils.

iv. develop allometric equations and estimate biomass and C stocks for

the five different forest types along the altitudinal gradient.


Page | 4

1.2 Thesis outline

The thesis is composed of four research chapters, preceded by a general

introduction and literature review, and finally a synthesis that includes a general

discussion. The first research chapter (Chapter 3) examines the effects of altitude and

forest composition on soil C and N stocks along the altitudinal gradient. In Chapter

4, stable isotopes in combination with elemental content of biomass and soil were

used to elucidate processes controlling C and N input, turnover and stability in the

soil for the different altitude forests. To holistically understand the SOC processes,

Chapter 5 examines the association of different forms of C in soil fractions of

varying density with soil minerals along the altitudinal gradient. In Chapter 6,

biomass equations were developed to estimate C stocks for the different forest types

along the altitudinal gradient. The final chapter of the thesis synthesises the main

finding of the research chapters and identifies some of the research gaps and future

directions for forest C stock estimations.


Page | 5

References

Achard, F., Beuchle, R., Mayaux, P., Stibig, H.-J., Bodart, C., Brink, A., Carboni, S.,

Desclée, B., Donnay, F., Eva, H.D., Lupi, A., Raši, R., Seliger, R., Simonetti,

D., 2014. Determination of tropical deforestation rates and related carbon

losses from 1990 to 2010. Global Change Biology 20(8), 2540-2554.

Bauhus, J., Paré, D., Co té, L., 1998. Effects of tree species, stand age and soil type

on soil microbial biomass and its activity in a southern boreal forest. Soil

Biology and Biochemistry 30(8–9), 1077-1089.

Chave, J., Andalo, C., Brown, S., Cairns, M., Chambers, J., Eamus, D., Fölster, H.,

Fromard, F., Higuchi, N., Kira, T., 2005. Tree allometry and improved

estimation of carbon stocks and balance in tropical forests. Oecologia 145(1),

87-99.

Chojnacky, D.C., Heath, L.S., Jenkins, J.C., 2014. Updated generalized biomass

equations for North American tree species. Forestry 87(1), 129-151.

Deng, X.W., Han, S.J., Hu, Y.L., Zhou, Y.M., 2009. Carbon and nitrogen

transformations in surface soils under Ermans Birch and dark coniferous

forests. Pedosphere 19(2), 230-237.

Dorji, T., Odeh, I.O.A., Field, D.J., Baillie, I.C., 2014. Digital soil mapping of soil

organic carbon stocks under different land use and land cover types in

montane ecosystems, Eastern Himalayas. Forest Ecology and Management

318, 91-102.

Eamus, D., McGuinness, K., Burrows, W., 2000. Review of allometric relationships

for estimating woody biomass for Queensland, the northern territory and

Western Australia. Queensland, Australia, Australian Greenhouse Office,

National Carbon Accounting System, Technical Report N5A.


Page | 6

Garkoti, S.C., 2008. Estimates of biomass and primary productivity in a high-altitude

maple forest of the west central Himalayas. Ecological Research 23(1), 41-

49.

Jenkins, J.C., Chojnacky, D.C., Heath, L.S., Birdsey, R.A., 2003. National-scale

biomass estimators for United States tree species. Forest Science 49(1), 12-

35.

Jobbagy, E.G., Jackson, R.B., 2000. The vertical distribution of soil organic carbon

and its relation to climate and vegetation. Ecological Applications 10(2), 423-

436.

Kieth, H., Barrett, D., Keenan, R., 1999. Review of allometric relationships for

estimating woody biomass for New South Wales, the Australian Capital

Territory, Victoria, Tasmania and South Australia.

Kindermann, G., Obersteiner, M., Sohngen, B., Sathaye, J., Andrasko, K.,

Rametsteiner, E., Schlamadinger, B., Wunder, S., Beach, R., 2008. Global

cost estimates of reducing carbon emissions through avoided deforestation.

Proceedings of the National Academy of Sciences 105(30), 10302-10307.

Lal, R., 2005. Forest soils and carbon sequestration. Forest Ecology and

Management 220(1), 242-258.

Le Quere, C., Raupach, M.R., Canadell, J.G., Marland, G., et al., 2009. Trends in the

sources and sinks of carbon dioxide. Nature Geoscience 2(12), 831-836.

Martin, D., Lal, T., Sachdev, C.B., Sharma, J.P., 2010. Soil organic carbon storage

changes with climate change, landform and land use conditions in Garhwal

hills of the Indian Himalayan mountains. Agriculture, Ecosystems and

Environment 138(1), 64-73.


Page | 7

Mendoza-Ponce, A., Galicia, L., 2010. Aboveground and belowground biomass and

carbon pools in highland temperate forest landscape in Central Mexico.

Forestry 83(5), 497-506.

Murty, D., Kirschbaum, M.U.F., McMurtrie, R.E., McGilvray, H., 2002. Does

conversion of forest to agricultural land change soil carbon and nitrogen? A

review of the literature. Global Change Biology 8(2), 105-123.

Muukkonen, P., 2007. Generalized allometric volume and biomass equations for

some tree species in Europe. European Journal of Forest Research 126(2),

157-166.

Muukkonen, P., Makipaa, R., Mencuccini, M., 2005. Biomass and stem volume

equations for tree species in Europe. Silva Fennica Monographs (4), 1-2,5-63.

Negi, K.S., Rawat, Y.S., Singh, J.S., 1983. Estimation of biomass and nutrient

storage in a Himalayan moist temperate forest. Canadian Journal of Forest

Research 13(6), 1185-1196.

Petersson, H., Holm, S., Ståhl, G., Alger, D., Fridman, J., Lehtonen, A., Lundström,

A., Mäkipää, R., 2012. Individual tree biomass equations or biomass

expansion factors for assessment of carbon stock changes in living biomass –

A comparative study. Forest Ecology and Management 270, 78-84.

Prichard, S.J., Peterson, D.L., Hammer, R.D., 2000. Carbon distribution in subalpine

forests and meadows of the Olympic mountains, Washington. Soil Science

Society of America Journal 64(5), 1834-1845.

Rana, B.S., Singh, S.P., Singh, R.P., 1989. Biomass and net primary productivity in

Central Himalayan forests along an altitudinal gradient. Forest Ecology and

Management 27(3–4), 199-218.


Page | 8

Singh, V., Tewari, A., Kushwaha, S.P.S., Dadhwal, V.K., 2011. Formulating

allometric equations for estimating biomass and carbon stock in small

diameter trees. Forest Ecology and Management 261(11), 1945-1949.

Snowdon, P., Eamus, D., Gibbons, P., Khanna, P., Keith, H., Raison, J., Kirschbaum,

M., 2000. Synthesis of allometrics, review of root biomass and design of

future woody biomass sampling strategies. Australian Greenhouse Office

Canberra, AU.

Somogyi, Z., Cienciala, E., Mäkipää, R., Muukkonen, P., Lehtonen, A., Weiss, P.,

2007. Indirect methods of large-scale forest biomass estimation. European

Journal of Forest Research 126(2), 197-207.

Stocker, T., Qin, D., Plattner, G., Tignor, M., Allen, S., Boschung, J., Nauels, A.,

Xia, Y., Bex, B., Midgley, B., 2013. IPCC, 2013: climate change 2013: the

physical science basis. Contribution of Working Group I to the Fifth

Assessment Report of the Intergovernmental Panel on Climate Change.

Ter-Mikaelian, M.T., Korzukhin, M.D., 1997. Biomass equations for sixty-five

North American tree species. Forest Ecology and Management 97(1), 1-24.

Zhang, M., Zhang, X.K., Liang, W.J., Jiang, Y., Dai, G.H., Wang, X.G., Han, S.J.,

2011. Distribution of soil organic carbon fractions along the altitudinal

gradient in Changbai Mountain, China. Pedosphere 21(5), 615-620.

Zianis, D., Muukkonen, P., Mäkipää, R., Mencuccini, M., 2005. Biomass and stem

volume equations for tree species in Europe. Silva Fennica Monographs, (4),

1–63.

Chapter 2. Literature Review

Page | 9

Chapter 2.

Literature Review

2.1 Forest carbon dynamics

The world at large is making concerted efforts to mitigate human-caused changes

in the Earth’s climate. International organisations, such as the United Nations

Framework Convention on Climate Change (UNFCC), bring countries together to

stabilize atmospheric greenhouse gases concentration. A scheme of incentives (debits

and credits) is under consideration to encourage specific changes in land use to

reduce the atmospheric concentration of carbon dioxide (Huston and Marland, 2003).

The atmospheric concentration of carbon dioxide has increased from a pre-industrial

value of 280 ppmv to 404.3 ppmv as of July 2016

(www.esrl.noaa.gov/gmd/ccgg/trends) and the trend indicates the rate continues to

rise unabated. The Mauna Loa observatory data show that the average increase in

CO2 emission rate that was 1.912 ppm for the decade from 1996 to 2005, has

increased to 2.170 ppm for the 2006 − 2015 decade. The most important human

induced sources of CO2 are from burning fossil fuel and the CO2 emission related to

land use change. Fossil fuel (e.g. oil or coal) buried deep inside the Earth is separated

from the normal terrestrial C cycle. However, when fossil fuels are extracted and

burnt as source of energy, CO2 is released into the atmosphere disturbing the natural

C balance. As CO2 traps heat in the atmosphere, increasing levels of CO2 in the

atmosphere alter the global climate. The emissions of CO2 from fossil fuel

combustion was 8.7 ± 0.5 Pg C year-1 in 2008, which is 29% greater than the value in

2000 (Le Quere et al., 2009). Consequently, the increase to global CO2 stock was 4.1

± 0.1 Pg C year-1 for the period 2000 to 2007, implying a balance of 4.6 Pg C per

year was absorbed by some other sinks like the ocean and forest (Pan et al., 2011).

A significant modern day challenge for humanity is to stop deforestation. If

forests ecosystems are left undisturbed, they have the potential to sequester and store


Page | 10

carbon (C) to mitigate climate change. However, forest degradation and destruction

are rampant. Forest cover declined at a rate of 1.0% per annum in Southeast Asia

from 2000 to 2010, equating to 11 M ha area of forest loss (Miettinen et al., 2011).

Another study, using Earth observation satellite data, net global forest loss was

estimated to 1.5 million square km between 2000 and 2012 period (Hansen et al.,

2013). The export of agricultural products on an industrial scale has been the driving

factor for forest destruction and land conversion in many areas (DeFries et al., 2010).

To mitigate net global C emissions effectively will require reduction in sources of

CO2 to the atmosphere as well as maintaining and increasing terrestrial and aquatic

sinks (Zhu, 2010). Apart from the atmosphere, other sinks of C, such as forests, soil

and oceans have evoked much interest in their ability to sequester C. Given the

importance of forests to act as a C sink, it is imperative to understand the current and

potential role of forest for international negotiations to limit greenhouse gas

emissions.

The Intergovernmental Panel on Climate Change (IPCC) estimated that terrestrial

ecosystems had a net uptake of C from 1.0 to 2.6 Pg C per year for 1990s (Nabuurs

and Karjalainen, 2007). However, a more recent study reported higher estimates of

2.0 to 3.4 Pg C per year (Pan et al., 2011). The UNFCCC has recognised the

importance of forests as a C sink as well as a source and requires countries to include

changes in forest C stocks in their annual greenhouse gas (GHG) inventories. Given

forest C balance is crucially linked to the atmosphere, the 7th Conference of Parties

(COP) to the UNFCCC agreed that countries under the Kyoto Protocol must account

for all forest C pools in their annual GHG inventories. Therefore, COP has

recognised the above and below ground biomass, deadwood, litter and soil organic C

as components of the forest C stock (Liski et al., 2006). The world’s forests are

estimated to contain up to 80% of all aboveground and 40% of belowground

terrestrial C (Dixon et al., 1994).


Page | 11

2.1.1 Soil organic carbon in terrestrial ecosystems

Globally, total soil organic carbon (SOC) content is estimated to be between

3,500 and 4,800 Pg C (Fig. 2.1) in the top 0 − 100 cm soil (Lehmann and Kleber,

2015). This makes SOC to be 5 − 10 times larger than the C stocks in the global

vegetation (Noble et al., 2000). The SOC pool in the top 100 cm soil can vary from

30 Mg C ha-1 to 800 Mg C ha-1 depending upon soil type and climatic conditions,

however most commonly, stocks range between 50 and 150 Mg C ha-1 (Lal, 2004).

Total global forest C stock is estimated to be 1146 Gt C and soil C constitutes about

69% of the total forest C stock (Dixon et al., 1994). The large quantity of C stored in

the soil therefore makes it an important sink or source of carbon dioxide (CO2),

depending upon factors such as climate, land use and land management practices.

Apart from sequestering C in the soils (Heimann and Reichstein, 2008), SOM also

retains nutrients and buffers pollutants which enhance plant growth and improves

water quality (Lal, 2004). The distribution of SOC is affected by numerous factors

including forest type (Deng et al., 2009), soil type (Bauhus et al., 1998), temperature

and precipitation (Jobbagy and Jackson, 2000; Zhang et al., 2011), as well as slope

and aspect (Dorji et al., 2015; Lal, 2005; Prichard et al., 2000). A study in the

Changbai mountains of China found no obvious altitudinal effect on SOC

distribution (Zhang et al., 2011). In contrast, in the Gharwal hills of India, at altitudes

greater than 1700 m the climate influence on C storage in soils was predominant over

vegetation type and landform, but at lower altitudes the vegetation type and

landforms effects were dominant over climate (Martin et al., 2010).

The SOC pool is in constant flux, with C cycling constantly between the C

reservoirs such as the atmosphere, biomass and oceans (Fig. 2.1). Plants through

photosynthesis remove CO2 from the atmosphere and store them as plant material.

Over time plants die, are burnt as fuel or decompose and release CO2 back to the

atmosphere. However, some of the organic matter (OM) is incorporated into the soil

as SOC. Soil organic carbon further decomposes releasing CO2 back to the

atmosphere or is bound to clay minerals and preserved in the soil for variable period

of time. Although the importance of SOC has long been recognized, the complexities


Page | 12

of the process involved with transformation of biomass into organic products and

their association with soil minerals make prediction of general trends of soil C

dynamics difficult. In order to consider soil C dynamics, it is important to understand

pool sizes, composition and their turnover times. Soil organic carbon is composed of

variable fractions based on functional pools. There is a small pool (1 – 5%) with a

rapid turnover time of a few weeks to months and two larger pools with slow

turnover rate from a few years to decades and very slow turnover rate of centuries

(Tirol-Padre and Ladha, 2004). Litter input, root biomass and microbial biomass

−responsible for litter and SOM transformation are the main C fractions

(Christensen, 1996). Litter inputs and SOM are subdivided into a readily

decomposable pool and a resistant pool of lignified materials (Hansen et al., 1991).

Microbial biomass in soil is separated as labile and physically protected pools

(Evans, 2001; Veen and Kuikman, 1990). As total SOC is composed of different C

fractions with varying degrees of stability (Jenkinson and Coleman, 1994; Veen et

al., 1984) total SOC as such may not adequately describe the role of C in many of the

soil processes (Janik et al., 2007).

Carbon turn over models generally consider the different C fractions and

decomposability of those fractions. However, the emerging view based on the soil

continuum model (SCM) focuses on protection of OM by the associated clay

minerals and its accessibility to the decomposers (Lehmann and Kleber, 2015). This

is in contrast to previous assumptions that SOM decomposition progressed via

preferential use of the more labile over more recalcitrant compounds (Vauramo and

Setälä, 2011). Under the emerging SCM concept, OM is continuously broken down

into smaller fragment by the decomposers. Smaller organic fragments have greater

opportunity to interact with mineral surfaces and to be incorporated into aggregates,

thereafter protects them from further decomposition (Lehmann and Kleber, 2015;

Lützow et al., 2006). The turnover times of mineral associated C on average were

shown to be four times longer than C in free or occluded OM, from a synthesis of

radiocarbon studies (Kleber et al., 2015).


Page | 13

Fig. 2.1 Total carbon stocks in different terrestrial components (i.e. soil, plant and

atmosphere) and annual fluxes (indicated by arrows) in and out of each of these

components (Lehmann and Kleber, 2015).

2.1.2 Preservation of soil organic carbon

Soil aggregate stability is a good indicator of soil quality affecting crop

production and soil sustainability (Amezketa, 1999). This is because soil aggregation

enhances soil properties such as porosity, hydraulic conductivity, water retention and

C stabilization (Cheng et al., 2015). Organo-mineral complexes, polysaccharides and

root exudates are the main organic constituents that stabilize soil aggregates (Arshad

and Coen, 1992). However, it is important to differentiate the intra-aggregate OM

that is incorporated and physically stabilized within macro-aggregates (Cambardella

and Elliott, 1992) and free OM found between soil aggregates (Six et al., 1998). The

location of SOM within the soil matrix determines the stability of the SOM by their

relative inaccessibility to decomposing soil organisms.

In an earlier work by Tisdall and Oades (1982), soil aggregates were physically

fractionated and classified into three main size classes, clay < 2 μm, micro-aggregate


Page | 14

(< 250 μm), and macro-aggregate (> 250 μm). For a more detailed classification, soil

aggregates have been further physically separated into large macro-aggregates (> 2

mm), small macro-aggregates or coarse inter-aggregate particulate organic matter

(250 − 2000 μm), micro-aggregates or fine inter-aggregate particulate organic matter

(250 − 53 μm) and the mineral fraction (< 53 μm) (Fernández-Ugalde et al., 2013;

Márquez et al., 2004). Different soil size class aggregate fractions are stabilised by

different mechanisms. The macro-aggregates > 250 μm are stabilised by plant roots

and hyphae, while for the micro-aggregates stabilisation depends upon the persistent

organic binding agents and soil characteristics (Tisdall and Oades, 1982). In essence,

the mechanisms of organo−mineral association are through physical, bio-chemical

and chemical stabilisation. Physical stabilisation is through preferential location of

organic matter in the soil structure (Six et al., 2002); biochemical stabilisation is

through the inherent characteristics of the OM to resist decomposition (Kögel‐

Knabner et al., 2008; Six et al., 2002) and chemical stabilisation through

intermolecular interactions between OM with clay minerals (Sollins et al., 1996) that

renders the OM inaccessible to decomposers. Soils with high C content had

substantive amount of C adsorbed onto mineral soil which lowers rates of

decomposition compared to soils with lower C content (Doetterl et al., 2015). These

differences in stabilisation mechanisms between macro and micro-aggregations, will

force them to respond differently to environmental factors and management practices

(Amezketa, 1999).

2.1.3 Characterisation of soil organic carbon

Sequential density fractionation of bulk soil with the use of sodium polytungstate

is employed to detect specific changes to different forms of C with change in

environmental factors and clay mineralogy (Diochon and Kellman, 2009; Six et al.,

2000). Fractionation of bulk soil through sequential density fractionation effectively

separates soil particles into particulate organic matter (POM) predominantly

mineral−free and increasingly organo−mineral particles with differing mineralogy

(Sollins et al., 2009). Soil organic carbon in various soil density fractions have


Page | 15

different C turnover periods as well as functions for C and nutrient dynamics (Crow

et al., 2007). Free POM reacts easily and decomposes relatively faster compared to

the more recalcitrant intra-aggregate matter and organo-mineral fraction (Poirier et

al., 2005). With increasing soil density fractions and increasing organo−mineral

association, SOC becomes increasingly stable (Baisden et al., 2002; Poirier et al.,

2005; Wagai et al., 2015).

To characterise SOC associated with minerals, the chemical pre-treatments which

were required to separate the OM from minerals can be avoided using spectroscopic

techniques (Lehmann and Kleber, 2015). Infrared radiation is passed through the

sample and based on the molecular structure of the sample; the radiation is adsorbed

or transmitted creating a unique spectrum. Thus, with the use of diffuse reflectance

infrared Fourier transform (DRIFT) spectroscopy, identification of different organic

functional groups and mineralogy in the soil samples can be achieved (Margenot et

al., 2015; Veum et al., 2014; Yeasmin et al., 2016). Combination of sequential

density fractionation of soil with DRIFT techniques to identify different SOC

associated with different clay mineralogy will help elucidate some of the complex

relationships that affect SOM stability in the soil.

2.1.4 Factors influencing SOC stocks

Soil organic carbon are dynamic in nature and influenced by environmental

factors, such as climate, vegetation (Maraseni and Pandey, 2014; Martin et al., 2010),

topography (Sharma et al., 2011) and soil texture (Jobbagy and Jackson, 2000). Soil

organic carbon in general has been found to be negatively correlated with annual

mean temperature and positively correlated with mean annual precipitation and

altitude (Dai and Huang, 2006; Lemenih and Itanna, 2004; Prietzel and Christophel,

2014). Even under similar biomass input, SOC content increased with increasing

altitude (Garten et al., 1999). This suggests that rate of OM decomposition driven by

altitudinal gradients of temperature is an important factor for altitudinal difference in

SOC stocks (Garten Jr and Hanson, 2006; Trumbore et al., 1996). In a recent study,

high altitude forest soils incubated at 15 °C over a 500 day period, revealed that only


Page | 16

the labile C pool was affected but not stable SOC stability (Tian et al., 2016). In

contrast, other soil incubation studies have reported a stronger response to increased

temperature for slowly decomposing C pools compared to more rapidly decomposing

SOC (Conant et al., 2008). Furthermore, δ13C in soils was reported to be negatively

correlated with temperature, implying that as temperature increased, recalcitrant C

was preferentially utilized (Biasi et al., 2005). However, fresh C input as source of

energy for the microbes needs consideration, as without it the stability of SOC

especially in deeper layers is maintained (Fontaine et al., 2007).

Biomass from vegetation is the main source of SOC. Different vegetation types,

such as forest, shrub lands, grasslands and croplands influence the vertical

distribution and SOC content in the soil. While vegetation and climate have a greater

influence on SOC at shallower depths, soil texture in more influential in determining

SOC content at deeper layers (Jobbagy and Jackson, 2000).

2.1.5 Forest biomass carbon

Current C stock in the world’s forest biomass is estimated to be 363 ± 28 Pg C,

which is about 42% of the total forest C stock (Pan et al., 2011). The forest biomass

is about 86% of the global vegetation (Sedjo, 1993), thereby making C stocks in

forest biomass an important C pool. From the global estimates by Dixon et al. (1994)

the C density for high, mid and low latitude forests were 64 Mg C ha-1, 57 Mg C ha-1

and 121 Mg C ha-1, respectively. A regional scale study in the cold temperate forest

of Veracruz central Mexico, on an altitudinal gradient from 2200 – 4000 m and

comprising seven forest types had aboveground biomass (AGB) C from 35.6 to

177.7 Mg C ha-1 (Mendoza-Ponce and Galicia, 2010). The C stocks varied

significantly between the forest types. From forest across northern China, Picea

−Abies forest and mixed conifer broadleaf forest had the highest mean biomass (178

– 202 Mg ha-1), while Pinus sylvestris forest had the lowest (78 Mg ha-1, Wang et al.,

2008). Similarly an estimate of the forest AGB ranged from 137 to 245 Mg ha-1 for

the Himalayan region of Uttar Pradesh in India (Haripriya, 2000), while Tiwari and

Singh (1987) estimated the AGB for India to range from 14 to 210 Mg ha-1. In the


Page | 17

North and Mid-Atlantic USA the maple-beech-birch forest had a mean aboveground

biomass of 97 Mg ha-1, while an oak-hickory forest had a mean AGB of 106 Mg ha-1

(Birdsey, 1992; Schroeder et al., 1997). The large variability in the AGB from

different forest types and regions is influenced by environmental factors such as,

precipitation, temperature, vegetation, topography, soil as well anthropogenic

management. In a forest eco-system apart from the SOC, various other pools such as

AGB, belowground root biomass, deadwood and litter components store a substantial

amount of C and needs to be estimated for accurate C accounting.

2.1.6 Measurement of the carbon pool in above ground tree biomass

Direct measurement of forest C stocks by felling and weighing is laborious,

expensive, time consuming and not practical. Therefore alternative methods like

aerial photogrammetry and light detection and ranging (LiDAR) techniques are

employed for forest biomass estimation. In the LiDAR method the estimation is

based on the reflectance information gathered, which has a relation to the forest

canopy, height and density. However, basic ecological information like tree height-

diameter, allometric and stand level wood density data has to be collected in the field

(Wulder et al., 2012). Additionally, LiDAR has been less often used due to large

data volumes and processing requirements (Skowronski and Lister, 2012).

Conventional forest inventory data for timber volume estimation are widely

collected and available throughout the world. The method of estimating biomass

stocks based on inventory data and C concentration to establish the corresponding C

stock is employed by many countries (Somogyi et al., 2007). However, most

conventional forest inventories have focused on timber volume and non-commercial

components were not considered. To account for the non-commercial components of

the forest biomass, biomass expansion factors (BEF) are used to estimate the total

forest biomass and C stock (Fang and Wang, 2001). By definition the BEF is the

ratio of above ground biomass to biomass of the merchantable timber. Still this

excludes smaller diameter understory vegetation. To account for the smaller stem

vegetation, the BEF has to be redefined to include the entire forest biomass


Page | 18

(Schroeder et al., 1997). Biomass expansion factors can vary with tree species and

size as the relative share of foliage and branches can vary with stand development

(Satoo and Madgwick, 1982). Biomass expansion factor is mostly reported as a

function of the stem volume (Schroeder et al., 1997), but it can be influenced by

other factors like branching pattern and sampling must be representative of the forest

population to reduce estimation errors (Somogyi et al., 2007).

A greater accuracy in estimating the forest biomass has been obtained using

individual tree representative biomass equations (Petersson et al., 2012). This implies

that stand- and species-specific biomass equations must be developed to reduce

estimation errors. A comprehensive database for species specific biomass and

volume equations are available for North America (Chojnacky et al., 2014; Ter-

Mikaelian and Korzukhin, 1997), Europe (Levy et al., 2004; Muukkonen, 2007) and

general biomass equations exist for pan tropical forest (Chave et al., 2005).

Biomass estimations in the Garwal Himalayas has been built on volume

equations developed earlier by the Forest Research Institute (FRI) and Forest Survey

of India (FSI), and as well as by multiplying by BEF developed by Brown and

Schroeder (1999) to convert the biomass volume to total AGB density (Gairola et al.,

2011). Biomass estimations using general allometric equations that ignored tree

species were found to be acceptable in a study in north-eastern China, but species-

specific equations particularly for branch and foliage biomass were recommended for

greater precision (Wang, 2006).

Measuring the AGB of an entire tree to develop biomass prediction models is

cumbersome and expensive. Randomized branch sampling (RBS), first adopted by

Jessen (1955), is a method for selecting sub-samples based on probability

proportional to size and is an efficient mean to estimate characteristics of trees, such

as aboveground woody dry matter (Valentine and Hilton, 1977) without having to

measure and weigh the entire tree. The estimator associated with RBS utilises inverse

probability weighting in a way that ensures unbiasedness (Gregoire et al., 1995).

Randomised branch sampling involves measuring the diameter of the bole at


Page | 19

successive fixed intervals from the base till a branch fork is encountered. The

diameter of each branch emanating from the fork is measured and a probability

proportional to its size is assigned. A random number is generated to randomly select

sampling pathway based on assigned probability proportional to bole and branch size

at each fork. The selected pathway needs to be marked and numbered to facilitate

data collection. The length of the bole or branch segment is measured. The procedure

is repeated at the next fork, and so on until a small branch or terminal shoot is

selected at the final node. This way the segments (sampling points) of the path

comprise a probability path for the entire tree or branch. Each segment of the

pathway is separated with loppers, and separated for woody and foliage biomass. The

weight of woody and foliage biomass of the segment are weighed. A sub sample is

further weighed, taken to the laboratory for oven drying and further analysis (Good

et al., 2001; Somogyi et al., 2007; Valentine et al., 1984). Oven dried biomass

subsamples are used to back calculate the biomass of the entire tree.

2.1.7 Carbon pool in below ground root biomass

Below ground biomass C stocks are difficult to measure by direct harvesting and

rarely have been measured for large areas. Rooting depths for vegetation across the

globe range from 29 cm for the tundra to 171 cm for Mediterranean shrub lands

(Schenk and Jackson, 2002), but the bulk of fine root biomass is generally found at

shallower depths, e.g. of less than 20 cm in a southern forest and as a consequence

only 8 − 16% of fine roots biomass was found at depths greater than 20 cm (Lukac

and Godbold, 2010).

Soil cores are generally extracted to estimate the root biomass. With this method,

a study from the highland temperate forest in central Mexico estimated root biomass

to be between 1.3 and 3.3% of the total biomass (Mendoza-Ponce and Galicia, 2010).

However, the estimates for below ground biomass based on soil cores or pits tend to

underestimate the biomass by not including the core root areas (Malhi et al., 2009).

Root biomass based on excavation of individual roots systems estimates the root

biomass to be between 17 and 28% for maple forest of west-central Himalayas


Page | 20

(Garkoti, 2008), and about 15% of the total tree biomass for dipterocarp forest in

Malaysia (Niiyama et al., 2010). Niiyama et al. (2010) stated that DBH of larger

trees (> 2.5 cm DBH) and collar diameter of smaller trees (< 2.5 cm DBH) are good

predictors of the coarse root biomass. Yet, the authors estimated that while

excavating, about 23% of the coarse roots were not recovered but needed to be

included for more accurate estimation of the total root biomass.

The root:shoot ratio is only a crude indicator of plant physiological processes that

determine the allocation of C, but it is vital for the estimation of belowground plant

biomass based on estimates for the aboveground plant biomass. In an analysis of

root:shoot for terrestrial biomes, this parameter was negatively correlated with

annual precipitation, mean annual temperature, forest stand age and stand height

(Mokany et al., 2006). The root:shoot ratio derived for specific regions and species

are used by national agencies to estimate belowground biomass from AGB which are

subsequently able to estimate the total biomass and C stocks for national greenhouse

gas inventory purposes (Cairns et al., 1997). From a synthesis of global surveys, the

root:shoot ratio for tropical regions was 0.21 ± 0.03 (Cairns et al., 1997; Jackson et

al., 1996). Although vegetation specific root:shoots ratio are expected to yield the

least error in estimating root biomass, many researchers have developed allometric

equations based on either the shoot biomass or DBH (Cairns et al., 1997; Vogt et al.,

1998; Vogt et al., 1996).

Based on relationships between the root biomass density (RBD) and root:shoot

ratios from available global data for temperate forests, the following equation was

developed to estimate RBD based on aboveground biomass density (AGBD) (Cairns

et al., 1997).

RBD= exp �– 1.0587 +0.8836 × ln (AGBD) + 0.2840� Equation 2.1

However, RBD along an altitudinal gradient from the subtropical to alpine

regions in Tibet was found to be correlated with temperature and precipitation rather

than the AGB. Root biomass density decreased with increasing altitude and had no


Page | 21

robust correlation to the AGB (Luo et al., 2005). These findings therefore suggest the

need for further investigation into RBD considering holistic parameters that are

correlated with physiological and environmental conditions and affect biomass

allocation to the above ground and below ground plant parts.

Although the need to estimate root biomass has been recognised given the

importance in estimating below ground C stocks, it is nonetheless difficult, laborious,

time consuming and ultimately not adequate as more that 23% of the coarse roots are

not recovered for measurement (Niiyama et al., 2010). The difficulty of physically

extracting root biomass for measurements ultimately limits the manageability of the

size and number of samples for adequate root biomass estimation. With

advancements in technology, various non-destructive root sampling methodologies

are being evaluated. Electrical resistivity tomography has had some success for

measuring root biomass density but not for root length density in an agricultural soil

(Rossi et al., 2011). Ground penetrating radar can measure root biomass to a limited

soil depth, but only when soil type and moisture conditions are ideal (Cui et al.,

2011; Cui et al., 2013).

2.1.8 Carbon pool in coarse woody debris

Dead wood (DW) is typically defined as all non-living woody tree biomass that is

standing or lying along with stumps (FAO 2006). As per the definition, DW may be

broken down to the following individual components: standing dead trees, down

dead wood (DDW), fine woody debris (FWD), stumps, and residue piles (Woodall et

al., 2009). The assessment of DW in forest ecosystems is required not only by the

interest in C accounting, but also for forest fire risk and biodiversity assessment.

Woody debris is influenced by vegetation, site conditions and natural and

anthropogenic disturbances. The deposition of woody debris is positively correlated

to the age of the stand (Fahey, 1983), and catastrophic events, such as hurricanes

(32.2 Mg ha-1) and forest fires (99.5 Mg ha-1) can quickly increase the woody debris

on the forest floor (Harmon et al., 1995). Deadwood store varies significantly with

succession stage and does not necessarily correlate with live biomass dynamics


Page | 22

(Krankina and Harmon, 1995). From a study along an elevation gradient in subalpine

Colorado, USA by Kueppers et al. (2004), the DW biomass increased by 75 kg ha-1

with every metre gain in elevation but decreased by 13 Mg ha-1 for every degree rise

in mean air temperature. This indicates that global warming can lead to a loss of

deadwood from subalpine forests.

Deadwood present in a forest will always be at some stage of decomposition. The

rate of decomposition is highly variable but in general quite slow due to low nutrient

content of woody tissues (Kueppers et al., 2004). Accurately measuring the decay

rate of wood is difficult due to its long turnover time. However as deadwood

decomposition progresses, density of the wood decreases (Paletto and Tosi, 2010)

and there is a slight increase in C content with increasing decay (Sandström et al.,

2007). Therefore a concurrent estimate of deadwood decay class is imperative for

reducing the estimation error while determining C stocks in the forest.

2.1.9 Carbon pool in coarse leaf litter

The IPCC guidelines define litter as an organic horizon (all leaves, twigs, small

branches, fruits, flowers, roots, and bark) on the mineral soil surface (IPCC., 2006).

In conifer forests, litter decomposes slowly due to high lignin content and cooler

temperatures (Green et al., 1993). The C contribution of leaf litter to SOC stock in a

deciduous forest was 6 – 9% and in coniferous forest between 17 and 49%

(Schrumpf et al., 2011). During decomposition of leaf litter between 60 and 80% of

the OC will be returned to the atmosphere, hence only a small portion either turns

into microbial biomass or forms humic substances after complex transformations

(González-Pérez et al., 2004). Carbon stocks for litter layer in the central highlands

in Mexico were found to be highly variable within the same forest and between

forest types. For fir forest, C stock ranged from 0.1 to 7.1 Mg C ha-1 with a mean of

4.1 Mg C ha-1; for the pine forest from 0.2 to 8.3 Mg C ha-1 with a mean of 3.0 Mg C

ha-1; and for the oak forest from 1.6 to 6.6 Mg C ha-1 with a mean of 3.2 Mg C ha-1

(Ordóñez et al., 2008). Carbon stock in litter may seem sizeable, but turnover time of

organic C in litter layer is only about three years (Schlesinger and Lichter, 2001).


Page | 23

2.1.10 Carbon content in biomass

Biomass is considered to contain 50% C and this value has been used to estimate

C stocks for tropical forests (Lewis et al., 2009; Saatchi et al., 2011), temperate

forests (Fang et al., 2001) and plantations (Beets et al., 2011). However, several

studies have found that the C concentration in biomass to vary between forest types

as well as between species and different tree components. Conifer trees had C

concentrations between 47.2 and 55.2%, while broadleaf trees had C concentrations

between 46.3 and 50.0% (Lamlom and Savidge, 2003; Thomas and Malczewski,

2007). A synthesis of C concentrations in tree tissues confirmed that conifers have a

greater C concentration than angiosperms (Thomas and Martin, 2012). Thus the

general accepted notion of 50% C concentration in the biomass needs to be revised,

which could reduce errors in the C content estimation of AGB by as much as 3.7%

(Thomas and Martin, 2012).

2.2 Summary

This review highlights factors that must be considered to estimate forest soil and

biomass C stocks and dynamics. For, instance the stability of SOC depends upon the

mechanisms of organo-mineral association, either through physical, bio-chemical or

through chemical stabilization. The mechanisms of organo-mineral association are in

turn dependent upon the clay content, type of clay minerals and the quantity and

quality of biomass input to the soil. However, during the physical, biological and

chemical transformation of organic matter and association with clay minerals,

complex processes are involved and many mechanisms of SOC and clay minerals are

yet to be understood. Climatic factors, such as precipitation and temperature, are

driving forces for SOC accumulation and decomposition and therefore should be

considered in relation to SOC dynamics. The monitoring of the SOC is imperative as

it is the largest proportion of the global terrestrial C stock and could further sequester

or release C depending upon anthropogenic activities.


Page | 24

The inputs of biomass to the soils are dependent upon the vegetation cover and

therefore are an essential component to monitor. The aboveground and below ground

biomass and C stocks especially in forest ecosystems have evoked a lot of interest

due to deforestation and conversion of forest land for agricultural purposes. Many

techniques and biomass equations are available for specific regions and species to

estimate their biomass. However, it has been established that local biomass equations

are necessary to achieve an acceptable level of accuracy of estimates. Most biomass

estimations have concentrated on the AGB, while few studies have been done in

relation to the belowground biomass component. Recovery of belowground biomass

is laborious and recovery rate of root biomass is low with high errors associated with

its estimate. Nonetheless, more effort needs to be made to increase the accuracy of C

stock estimates of forests and understanding of its dynamics. This is especially

important as the success of the international programmes such as United Nations

Reducing Emissions from Deforestation and Forest Degradation (UN-REDD)

program to reduce forest C emissions and enhance C stocks hinges on robust and

comprehensive C stock estimations and monitoring processes.


Page | 25

References

Amezketa, E., 1999. Soil aggregate stability: a review. Journal of Sustainable

Agriculture 14(2-3), 83-151.

Arshad, M., Coen, G., 1992. Characterization of soil quality: physical and chemical

criteria. American Journal of Alternative Agriculture 7(1-2), 25-31.

Baisden, W., Amundson, R., Cook, A., Brenner, D., 2002. Turnover and storage of C

and N in five density fractions from California annual grassland surface soils.

Global Biogeochemical Cycles 16(4).

Bauhus, J., Paré, D., Co té, L., 1998. Effects of tree species, stand age and soil type

on soil microbial biomass and its activity in a southern boreal forest. Soil

Biology and Biochemistry 30(8–9), 1077-1089.

Beets, P.N., Brandon, A.M., Goulding, C.J., Kimberley, M.O., Paul, T.S.H., Searles,

N., 2011. The inventory of carbon stock in New Zealand’s post-1989 planted

forest for reporting under the Kyoto protocol. Forest Ecology and

Management 262(6), 1119-1130.

Biasi, C., Rusalimova, O., Meyer, H., Kaiser, C., Wanek, W., Barsukov, P., Junger,

H., Richter, A., 2005. Temperature-dependent shift from labile to recalcitrant

carbon sources of arctic heterotrophs. Rapid Communications in Mass

Spectrometry 19(11), 1401-1408.

Birdsey, R.A., 1992. Carbon storage and accumulation in United States forest

ecosystems. (Gen Tech Rep WO-59). USDA Forest Service, Washington DC.

Brown, S.L., Schroeder, P.E., 1999. Spatial patterns of aboveground production and

mortality of woody biomass for Eastern U.S. forest. Ecological Applications

9(3), 968-980.

Cairns, M.A., Brown, S., Helmer, E.H., Baumgardner, G.A., 1997. Root biomass

allocation in the world's upland forests. Oecologia 111(1), 1-11.


Page | 26

Cambardella, C.A., Elliott, E.T., 1992. Particulate soil organic matter changes across

a grassland cultivation sequence. Soil Science Society of America Journal

56(3), 777-783.




87-99.

Cheng, M., Xiang, Y., Xue, Z., An, S., Darboux, F., 2015. Soil aggregation and intra-

aggregate carbon fractions in relation to vegetation succession on the Loess

Plateau, China. Catena 124, 77-84.



Christensen, B., 1996. Matching measurable soil organic matter fractions with

conceptual pools in simulation models of carbon turnover: Revision of model

structure. In: D. Powlson, P. Smith, J. Smith (Eds.), Evaluation of Soil

Organic Matter Models. NATO ASI Series. Springer Berlin Heidelberg, pp.

143-159.

Conant, R.T., Drijber, R.A., Haddix, M.L., Parton, W.J., Paul, E.A., Plante, A.F.,

Six, J., Steinweg, J.M., 2008. Sensitivity of organic matter decomposition to

warming varies with its quality. Global Change Biology 14(4), 868-877.

Crow, S.E., Swanston, C.W., Lajtha, K., Brooks, J.R., Keirstead, H., 2007. Density

fractionation of forest soils: Methodological questions and interpretation of

incubation results and turnover time in an ecosystem context.

Biogeochemistry 85(1), 69-90.

Cui, X., Chen, J., Shen, J., Cao, X., Chen, X., Zhu, X., 2011. Modeling tree root

diameter and biomass by ground-penetrating radar. Science China Earth

Sciences 54(5), 711-719.


Page | 27

Cui, X., Guo, L., Chen, J., Chen, X., Zhu, X., 2013. Estimating tree-root biomass in

different depths using ground-penetrating radar: evidence from a controlled

experiment. Geoscience and Remote Sensing, IEEE Transactions on 51(6),

3410-3423.

Dai, W., Huang, Y., 2006. Relation of soil organic matter concentration to climate

and altitude in zonal soils of China. Catena 65(1), 87-94.

DeFries, R.S., Rudel, T., Uriarte, M., Hansen, M., 2010. Deforestation driven by

urban population growth and agricultural trade in the twenty-first century.

Nature Geoscience 3(3), 178-181.

Deng, X.W., Han, S.J., Hu, Y.L., Zhou, Y.M., 2009. Carbon and nitrogen

transformations in surface soils under Ermans Birch and dark coniferous

forests. Pedosphere 19(2), 230-237.

Diochon, A.C., Kellman, L., 2009. Physical fractionation of soil organic matter:

Destabilization of deep soil carbon following harvesting of a temperate

coniferous forest. Journal of Geophysical Research: Biogeosciences 114(G1),

1 - 9.

Dixon, R.K., Houghto, R.A., Solomon, A.M., Trexler, M.C., Wisniewski, J., 1994.

Carbon pools and flux of global forest ecosystems. Science 263, 185-190.

Doetterl, S., Stevens, A., Six, J., Merckx, R., Van Oost, K., Pinto, M.C., Casanova-

Katny, A., Muñoz, C., Boudin, M., Venegas, E.Z., 2015. Soil carbon storage

controlled by interactions between geochemistry and climate. Nature

Geoscience 8(10), 780-783.

Dorji, T., Odeh, I.A., Field, D., 2015. Elucidating the complex interrelationships of

soil organic carbon fractions with land use/land cover types and landform

attributes in a montane ecosystem. Soils and Sediments 15(5), 1039-1054.


Page | 28

Evans, J.L., Fernandez, I. J., Rustad, L. E., Norton, S. A., 2001. Methods for

evaluating carbon fractions in forest soils: a review. Technical Bulletin 178.

Fahey, T.J., 1983. Nutrient dynamics of aboveground detritus in lodgepole pine

(Pinus contorta ssp. latifolia) ecosystems, Southeastern Wyoming. Ecological

Monographs 53(1), 51-72.

Fang, J.-Y., Wang, Z.M., 2001. Forest biomass estimation at regional and global

levels, with special reference to China's forest biomass. Ecological Research

16(3), 587-592.

Fang, J., Chen, A., Peng, C., Zhao, S., Ci, L., 2001. Changes in forest biomass

carbonstorage in China between 1949 and 1998. Science 292(5525), 2320-

2322.

Fernández-Ugalde, O., Barré, P., Hubert, F., Virto, I., Girardin, C., Ferrage, E.,

Caner, L., Chenu, C., 2013. Clay mineralogy differs qualitatively in

aggregate-size classes: clay-mineral-based evidence for aggregate hierarchy

in temperate soils. European Journal of Soil Science 64(4), 410-422.

Fontaine, S., Barot, S., Barre, P., Bdioui, N., Mary, B., Rumpel, C., 2007. Stability of

organic carbon in deep soil layers controlled by fresh carbon supply. Nature

450(7167), 277-280.

Gairola, S., Sharma, C.M., Ghildiyal, S.K., Suyal, S., 2011. Live tree biomass and

carbon variation along an altitudinal gradient in moist temperate valley slopes

of the Garhwal Himalaya (India). Current Science 100(12), 1862-1870.



49.


Page | 29

Garten, C.T., Jr., Post, W.M., III, Hanson, P.J., Cooper, L.W., 1999. Forest soil

carbon inventories and dynamics along an elevation gradient in the southern

Appalachian Mountains. Biogeochemistry 45(2), 115-145.

Garten Jr, C.T., Hanson, P.J., 2006. Measured forest soil C stocks and estimated

turnover times along an elevation gradient. Geoderma 136(1–2), 342-352.

González-Pérez, J.A., González-Vila, F.J., Almendros, G., Knicker, H., 2004. The

effect of fire on soil organic matter—a review. Environment International

30(6), 855-870.

Good, N., Paterson, M., Brack, C., Mengersen, K., 2001. Estimating tree component

biomass using variable probability sampling methods. JABES 6(2), 258-267.

Green, R.N., Trowbridge, R.L., Klinka, K., 1993. Towards a taxonomic classification

of humus forms. Forest Science 39(Supplement 28), 1-2.

Gregoire, T.G., Valentine, H.T., Furnival, G.M., 1995. Sampling methods to estimate

foliage and other characteristics of individual trees. Ecology 76(4), 1181-

1194.

Hansen, M.C., Potapov, P.V., Moore, R., Hancher, M., Turubanova, S., Tyukavina,

A., Thau, D., Stehman, S., Goetz, S., Loveland, T., 2013. High-resolution

global maps of 21st-century forest cover change. Science 342(6160), 850-

853.

Hansen, S., Jensen, H.E., Nielsen, N.E., Svendsen, H., 1991. Simulation of nitrogen

dynamics and biomass production in winter wheat using the Danish

simulation model DAISY. Fertilizer Research 27(2-3), 245-259.

Haripriya, G.S., 2000. Estimates of biomass in Indian forests. Biomass and

Bioenergy 19(4), 245-258.


Page | 30

Harmon, M.E., Whigham, D.F., Sexton, J., Olmsted, I., 1995. Decomposition and

mass of woody detritus in the dry tropical forests of the Northeastern Yucatan

Peninsula, Mexico. Biotropica 27(3), 305-316.

Heimann, M., Reichstein, M., 2008. Terrestrial ecosystem carbon dynamics and

climate feedbacks. Nature 451(7176), 289-292.

Huston, M.A., Marland, G., 2003. Carbon management and biodiversity. Journal of

Environmental Management 67(1), 77-86.

IPCC., 2006. IPCC guidelines for national greenhouse gas inventories. Institute for

global environmental strategies (IGES) for the IPCC. Kanagawa, Japan.

Jackson, R.B., Canadell, J., Ehleringer, J.R., Mooney, H.A., Sala, O.E., Schulze,

E.D., 1996. A global analysis of root distributions for terrestrial biomes.

Oecologia 108(3), 389-411.

Janik, L.J., Skjemstad, J.O., Shepherd, K.D., Spouncer, L.R., 2007. The prediction of

soil carbon fractions using mid-infrared-partial least square analysis. Soil

Research 45(2), 73-81.

Jenkinson, D.S., Coleman, K., 1994. Calculating the annual input of organic matter

to soil from measurements of total organic carbon and radiocarbon. European

Journal of Soil Science 45(2), 167-174.

Jessen, R.J., 1955. Determining the fruit count on a tree by randomized branch

sampling. Biometrics 11(1), 99-109.



436.

Kleber, M., Eusterhues, K., Keiluweit, M., Mikutta, C., Mikutta, R., Nico, P.S.,

2015. Mineral–organic associations: Formation, properties, and relevance in

soil environments, Advances in Agronomy, pp. 1-140.


Page | 31

Kögel‐Knabner, I., Guggenberger, G., Kleber, M., Kandeler, E., Kalbitz, K., Scheu,

S., Eusterhues, K., Leinweber, P., 2008. Organo‐mineral associations in

temperate soils: Integrating biology, mineralogy, and organic matter

chemistry. Journal of Plant Nutrition and Soil Science 171(1), 61-82.

Krankina, O.N., Harmon, M.E., 1995. Dynamics of the dead wood carbon pool in

Northwestern Russian boreal forests. In: M. Apps, D. Price, J. Wisniewski

(Eds.), Boreal Forests and Global Change. Springer Netherlands, pp. 227-

238.

Kueppers, L., Southon, J., Baer, P., Harte, J., 2004. Dead wood biomass and turnover

time, measured by radiocarbon, along a subalpine elevation gradient.

Oecologia 141(4), 641-651.

Lal, R., 2004. Soil carbon sequestration impacts on global climate change and food

security. Science 304(5677), 1623-1627.

Lal, R., 2005. Forest soils and carbon sequestration. Forest Ecology and

Management 220(1), 242-258.

Lamlom, S.H., Savidge, R.A., 2003. A reassessment of carbon content in wood:

variation within and between 41 North American species. Biomass and

Bioenergy 25(4), 381-388.

Le Quere, C., Raupach, M.R., Canadell, J.G., Marland, G., et al., 2009. Trends in the

sources and sinks of carbon dioxide. Nature Geoscience 2(12), 831-836.

Lehmann, J., Kleber, M., 2015. The contentious nature of soil organic matter. Nature

528(7580), 60-68.

Lemenih, M., Itanna, F., 2004. Soil carbon stocks and turnovers in various vegetation

types and arable lands along an elevation gradient in southern Ethiopia.

Geoderma 123(1–2), 177-188.


Page | 32

Levy, P.E., Hale, S.E., Nicoll, B.C., 2004. Biomass expansion factors and root :

shoot ratios for coniferous tree species in Great Britain. Forestry 77(5), 421-

430.

Lewis, S.L., Lopez-Gonzalez, G., Sonké, B., Affum-Baffoe, K., Baker, T.R., Ojo,

L.O., Phillips, O.L., Reitsma, J.M., White, L., Comiskey, J.A., 2009.

Increasing carbon storage in intact African tropical forests. Nature 457(7232),

1003-1006.

Liski, J., Lehtonen, A., Palosuo, T., Peltoniemi, M., Eggers, T., Muukkonen, P.,

Mäkipää, R., 2006. Carbon accumulation in Finland's forests 1922–2004 – an

estimate obtained by combination of forest inventory data with modelling of

biomass, litter and soil. Annals of Forest Science. 63(7), 687-697.

Lukac, M., Godbold, D., 2010. Fine root biomass and turnover in southern taiga

estimated by root inclusion nets. Plant and Soil 331(1-2), 505-513.

Luo, T., Brown, S., Pan, Y., Shi, P., Ouyang, H., Yu, Z., Zhu, H., 2005. Root

biomass along subtropical to alpine gradients: global implication from

Tibetan transect studies. Forest Ecology and Management 206(1), 349-363.

Lützow, M.V., Kögel‐Knabner, I., Ekschmitt, K., Matzner, E., Guggenberger, G.,

Marschner, B., Flessa, H., 2006. Stabilization of organic matter in temperate

soils: mechanisms and their relevance under different soil conditions–a

review. European Journal of Soil Science 57(4), 426-445.

Malhi, Y., AragÃO, L.E.O.C., Metcalfe, D.B., Paiva, R., Quesada, C.A., Almeida,

S., Anderson, L., Brando, P., Chambers, J.Q., Da Costa, A.C.L., Hutyra, L.R.,

Oliveira, P., PatiÑO, S., Pyle, E.H., Robertson, A.L., Teixeira, L.M., 2009.

Comprehensive assessment of carbon productivity, allocation and storage in

three Amazonian forests. Global Change Biology 15(5), 1255-1274.


Page | 33

Maraseni, T.N., Pandey, S.S., 2014. Can vegetation types work as an indicator of soil

organic carbon? An insight from native vegetations in Nepal. Ecological

Indicators 46(0), 315-322.

Margenot, A.J., Calderón, F.J., Bowles, T.M., Parikh, S.J., Jackson, L.E., 2015. Soil

organic matter functional group composition in relation to organic carbon,

nitrogen, and phosphorus fractions in organically managed tomato fields. Soil

Science Society of America Journal 79(3), 772-782.

Márquez, C., Garcia, V., Cambardella, C., Schultz, R., Isenhart, T., 2004. Aggregate-

size stability distribution and soil stability. Soil Science Society of America

Journal 68(3), 725-735.

Martin, D., Lal, T., Sachdev, C.B., Sharma, J.P., 2010. Soil organic carbon storage

changes with climate change, landform and land use conditions in Garhwal

hills of the Indian Himalayan mountains. Agriculture, Ecosystems and

Environment 138(1), 64-73.

Mendoza-Ponce, A., Galicia, L., 2010. Aboveground and belowground biomass and

carbon pools in highland temperate forest landscape in Central Mexico.

Forestry 83(5), 497-506.

Miettinen, J., Shi, C., Liew, S.C., 2011. Deforestation rates in insular Southeast Asia

between 2000 and 2010. Global Change Biology 17(7), 2261-2270.

Mokany, K., Raison, R.J., Prokushkin, A.S., 2006. Critical analysis of root : shoot

ratios in terrestrial biomes. Global Change Biology 12(1), 84-96.



157-166.

Nabuurs, G.J., O. Masera, K. Andrasko, P. Benitez-Ponce, R. Boer, M. Dutschke, E.

Elsiddig, J. Ford-Robertson, P. Frumhoff, T., Karjalainen, O.K., W.A. Kurz,


Page | 34

M. Matsumoto, W. Oyhantcabal, N.H. Ravindranath, M.J. Sanz Sanchez, X.

Zhang, 2007. Forestry In Climate Change 2007: Mitigation. Contribution of

Working Group III to the Fourth Assessment Report of the Intergovernmental

Panel on Climate Change, Cambridge University Press, Cambridge, United

Kingdom and New York, NY, USA.

Niiyama, K., Kajimoto, T., Matsuura, Y., Yamashita, T., Matsuo, N., Yashiro, Y.,

Ripin, A., Kassim, A.R., Noor, N.S., 2010. Estimation of root biomass based

on excavation of individual root systems in a primary dipterocarp forest in

Pasoh Forest Reserve, Peninsular Malaysia. Journal of Tropical Ecology

26(3), 271-284.

Noble, I., Bolin, B., Ravindranath, N., Verardo, D., Dokken, D., (Ed), 2000. Land

Use, Land‐Use Change, and Forestry. Cambridge University Press.

Ordóñez, J.A.B., de Jong, B.H.J., García-Oliva, F., Aviña, F.L., Pérez, J.V.,

Guerrero, G., Martínez, R., Masera, O., 2008. Carbon content in vegetation,

litter, and soil under 10 different land-use and land-cover classes in the

Central Highlands of Michoacan, Mexico. Forest Ecology and Management

255(7), 2074-2084.

Paletto, A., Tosi, V., 2010. Deadwood density variation with decay class in seven

tree species of the Italian Alps. Scandinavian Journal of Forest Research

25(2), 164-173.

Pan, Y., Birdsey, R.A., Fang, J., Houghton, R., Kauppi, P.E., Kurz, W.A., Phillips,

O.L., Shvidenko, A., Lewis, S.L., Canadell, J.G., Ciais, P., Jackson, R.B.,

Pacala, S.W., McGuire, A.D., Piao, S., Rautiainen, A., Sitch, S., Hayes, D.,

2011. A large and persistent carbon sink in the world’s forests. Science

333(6045), 988-993.




Page | 35



Poirier, N., Sohi, S.P., Gaunt, J.L., Mahieu, N., Randall, E.W., Powlson, D.S.,

Evershed, R.P., 2005. The chemical composition of measurable soil organic

matter pools. Organic Geochemistry 36(8), 1174-1189.

Prichard, S.J., Peterson, D.L., Hammer, R.D., 2000. Carbon distribution in subalpine

forests and meadows of the Olympic mountains, Washington. Soil Science


Prietzel, J., Christophel, D., 2014. Organic carbon stocks in forest soils of the

German Alps. Geoderma 221–222(0), 28-39.

Rossi, R., Amato, M., Bitella, G., Bochicchio, R., Ferreira Gomes, J.J., Lovelli, S.,

Martorella, E., Favale, P., 2011. Electrical resistivity tomography as a non-

destructive method for mapping root biomass in an orchard. European Journal

of Soil Science 62(2), 206-215.

Saatchi, S.S., Harris, N.L., Brown, S., Lefsky, M., Mitchard, E.T.A., Salas, W.,

Zutta, B.R., Buermann, W., Lewis, S.L., Hagen, S., Petrova, S., White, L.,

Silman, M., Morel, A., 2011. Benchmark map of forest carbon stocks in

tropical regions across three continents. Proceedings of the National

Academy of Sciences 108(24), 9899-9904.

Sandström, F., Petersson, H., Kruys, N., Ståhl, G., 2007. Biomass conversion factors

(density and carbon concentration) by decay classes for dead wood of Pinus

sylvestris, Picea abies and Betula spp. in boreal forests of Sweden. Forest

Ecology and Management 243(1), 19-27.

Satoo, T., Madgwick, H.A.I., 1982. Forest biomass. Forestry Sciences. Nijhoff The

Hague, 152 pp.


Page | 36

Schenk, H.J., Jackson, R.B., 2002. The Global biogeography of roots. Ecological

Monographs 72(3), 311-328.

Schlesinger, W.H., Lichter, J., 2001. Limited carbon storage in soil and litter of

experimental forest plots under increased atmospheric CO2. Nature

411(6836), 466-469.

Schroeder, P., Brown, S., Mo, J., Birdsey, R., Cieszewski, C., 1997. Biomass

estimation for temperate broadleaf forest of the United States using inventory

data. Forest Science 43(3), 424-434.

Schrumpf, M., Schulze, E.D., Kaiser, K., Schumacher, J., 2011. How accurately can

soil organic carbon stocks and stock changes be quantified by soil

inventories? Biogeosciences 8(5), 1193-1212.

Sedjo, R., 1993. The carbon cycle and global forest ecosystem. Water, Air, and Soil

Pollution 70(1-4), 295-307.

Sharma, C.M., Gairola, S., Baduni, N.P., Ghildiyal, S.K., Suyal, S., 2011. Variation

in carbon stocks on different slope aspects in seven major forest types of

temperate region of Garhwal Himalaya, India. Journal of Biosciences 36(4),

701-708.

Six, J., Conant, R.T., Paul, E.A., Paustian, K., 2002. Stabilization mechanisms of soil

organic matter: Implications for C-saturation of soils. Plant and Soil 241(2),

155-176.

Six, J., Elliott, E., Paustian, K., Doran, J., 1998. Aggregation and soil organic matter

accumulation in cultivated and native grassland soils. Soil Science Society of

America Journal 62(5), 1367-1377.

Six, J., Merckx, R., Kimpe, K., Paustian, K., Elliott, E., 2000. A re‐evaluation of the

enriched labile soil organic matter fraction. European Journal of Soil Science

51(2), 283-293.


Page | 37

Skowronski, N.S., Lister, A.J., 2012. Utility of LiDAR for large area forest inventory

applications, Moving from Status to Trends: Forest Inventory and Analysis

(FIA) Symposium, pp. 410-413.

Sollins, P., Homann, P., Caldwell, B.A., 1996. Stabilization and destabilization of

soil organic matter: mechanisms and controls. Geoderma 74(1), 65-105.

Sollins, P., Kramer, M.G., Swanston, C., Lajtha, K., Filley, T., Aufdenkampe, A.K.,

Wagai, R., Bowden, R.D., 2009. Sequential density fractionation across soils

of contrasting mineralogy: evidence for both microbial- and mineral-

controlled soil organic matter stabilization. Biogeochemistry 96(1), 209-231.






Thomas, S.C., Malczewski, G., 2007. Wood carbon content of tree species in Eastern

China: Interspecific variability and the importance of the volatile fraction.

Journal of Environmental Management 85(3), 659-662.

Thomas, S.C., Martin, A.R., 2012. Carbon content of tree tissues: A synthesis.

Forests 3(2), 332-352.

Tian, Q., He, H., Cheng, W., Bai, Z., Wang, Y., Zhang, X., 2016. Factors controlling

soil organic carbon stability along a temperate forest altitudinal gradient.

Scientific Reports 6, 18783.

Tirol-Padre, A., Ladha, J.K., 2004. Assessing the reliability of permanganate-

oxidizable carbon as an index of soil labile carbon. Soil Science Society



Page | 38

Tisdall, J.M., Oades, J.M., 1982. Organic matter and water-stable aggregates in soils.

Journal of Soil Science 33(2), 141-163.

Tiwari, A.K., Singh, J.S., 1987. Analysis of forest land-use and vegetation in a part

of Central Himalaya, using aerial photographs. Environmental Conservation

14(03), 233-244.

Trumbore, S.E., Chadwick, O.A., Amundson, R., 1996. Rapid exchange between soil

carbon andatmospheric carbon dioxide driven by temperature change.

Science 272(5260), 393-396.

Valentine, H.T., Hilton, S.J., 1977. Sampling oak foliage by the randomized-branch

method. Canadian Journal of Forest Research 7(2), 295-298.

Valentine, H.T., Tritton, L.M., Furnival, G.M., 1984. Subsampling trees for biomass,

volume, or mineral content. Forest Science 30(3), 673-681.

Vauramo, S., Setälä, H., 2011. Decomposition of labile and recalcitrant litter types

under different plant communities in urban soils. Urban Ecosystems 14(1),

59-70.

Veen, J.A., Kuikman, P.J., 1990. Soil structural aspects of decomposition of organic

matter by micro-organisms. Biogeochemistry 11(3), 213-233.

Veen, J.A., Ladd, J.N., Frissel, M.J., 1984. Modelling C and N turnover through the

microbial biomass in soil. Plant and Soil 76(1-3), 257-274.

Veum, K.S., Goyne, K.W., Kremer, R.J., Miles, R.J., Sudduth, K.A., 2014.

Biological indicators of soil quality and soil organic matter characteristics in

an agricultural management continuum. Biogeochemistry 117(1), 81-99.

Vogt, K., Vogt, D., Bloomfield, J., 1998. Analysis of some direct and indirect

methods for estimating root biomass and production of forests at an

ecosystem level. Plant and Soil 200(1), 71-89.


Page | 39

Vogt, K.A., Vogt, D.J., Palmiotto, P.A., Boon, P., O'Hara, J., Asbjornsen, H., 1996.

Review of root dynamics in forest ecosystems grouped by climate, climatic

forest type and species. Plant and Soil 187(2), 159-219.

Wagai, R., Kajiura, M., Asano, M., Hiradate, S., 2015. Nature of soil organo-mineral

assemblage examined by sequential density fractionation with and without

sonication: Is allophanic soil different? Geoderma 241, 295-305.

Wang, C., 2006. Biomass allometric equations for 10 co-occurring tree species in

Chinese temperate forests. Forest Ecology and Management 222(1), 9-16.

Wang, X., Fang, J., Zhu, B., 2008. Forest biomass and root–shoot allocation in

northeast China. Forest Ecology and Management 255(12), 4007-4020.

Woodall, C., Rondeux, J., Verkerk, P., Ståhl, G., 2009. Estimating dead wood during

national forest inventories: A review of inventory methodologies and

suggestions for harmonization. Environmental Management 44(4), 624-631.

Wulder, M.A., White, J.C., Nelson, R.F., Naesset, E., Ørka, H.O., Coops, N.C.,

Hilker, T., Bater, C.W., Gobakken, T., 2012. Lidar sampling for large-area

forest characterization: A review. Remote Sensing of Environment 121, 196-

209.

Yeasmin, S., Singh, B., Johnston, C.T., Sparks, D.L., 2016. Evaluation of pre-

treatment procedures for improved interpretation of mid infrared spectra of

soil organic matter. Geoderma.

Zhang, M., Zhang, X.K., Liang, W.J., Jiang, Y., Dai, G.H., Wang, X.G., Han, S.J.,

2011. Distribution of soil organic carbon fractions along the altitudinal

gradient in Changbai Mountain, China. Pedosphere 21(5), 615-620.

Zhu, Z., ed., Bergamaschi, Brian, Bernknopf, Richard, Clow, David, Dye, Dennis,

Faulkner, Stephen, Forney, Wil-liam, Gleason, Robert, Hawbaker, Todd, Liu,

Jinxun, Liu, Shuguang, Prisley, Stephen, Reed, Bradley, Reeves, Matthew,


Page | 40

Rollins, Matthew, Sleeter, Benjamin, Sohl, Terry, Stackpoole, Sarah,

Stehman, Stephen, Striegl, Robert, Wein, Anne, and Zhu, Zhiliang, 2010.

Public review draft: A method for assessing carbon stocks, carbon

sequestration, and greenhouse-gas fluxes in ecosystems of the United States

under present conditions and future scenarios: U.S. Geological Survey Open-

File Report 2010-1144

Chapter 3. Soil carbon and nitrogen stocks

Page | 41

Chapter 3.

Soil carbon and nitrogen stocks in forests along an

altitudinal gradient in the eastern Himalayas and a

meta-analysis of global data1

Abstract

High altitude soils potentially store a large pool of carbon (C) and nitrogen

(N).The assessment of total C and N stocks in soils is vital to understanding the C

and N dynamics in terrestrial ecosystems. In this study we examined effects of

altitude and forest composition on soil C and N along a transect from 317 to 3300 m

a.s.l. in the eastern Himalayas. We used meta-analysis to establish the context for

our results on the effects of altitude on soil C, including variation with depth. Total C

and N content of soils significantly increased with altitude, but decreased with soil

depth. Carbon and N were similarly correlated with altitude and temperature; and

temperature was seemingly the main driver of soil C along the altitudinal gradient.

Altitude accounted for 73% of the variation in C and 47% of the variation in N

stocks. Soil pH and cat-ion exchange capacity (CEC) were correlated with both soil

C and N stocks. Increasing soil C and N stocks were related to forest composition,

forest basal area (BA) as well as quantity of leaf litter that were in turn influenced by

altitude and temperature. Concentrations of C in foliage increased by 2.1% for every

1000 m rise in altitude, while that in leaf litter increased by 2.3%.

1 This chapter has been published in Global Change Biology under the title “Soil carbon and nitrogen stocks in forest along an altitudinal gradient in the eastern Himalayas and a meta-analysis of global data” in January 2016. Authors are Sonam Tashi, Balwant Singh, Claudia Keitel and Mark Adams


Page | 42

3.1 Introduction

Carbon stocks in soil vary substantially across the globe depending on the type

of forest, their location and soil depth. In mixed forests C stocks are highly variable,

with 63 – 88 Mg C ha-1 reported for Picea and Abies, mixed broadleaf and Betula

forests in the northeast of China (Zhu et al., 2010); 184 Mg C ha-1 (0 – 100 cm) for a

Picea dominated forest in Poland (Galka et al., 2014); and 93 – 101 Mg C ha-1 (0 – 30

cm) for an Abies and pine-oak forest in Mexico (Ordóñez et al., 2008). Similarly, highly

variable soil C stocks (0 – 100 cm) have been reported in a study from India, with

values ranging from 34 to 411 Mg ha-1 for tropical evergreen forest, from 24 to 525 Mg

ha-1 for montane temperate forest, and from 57 to 213 Mg ha-1 for tropical moist

deciduous forest (Chhabra et al., 2003). These highly variable data for soil C stocks for

different forest types across different geographical zones highlight the importance of

site-specific measurements of parameters with predictive capacity for a good

approximation of C stocks in different ecosystems around the world.

Limited data are available for the N stock for natural forests in relation to

forest type or along altitudinal gradients. However, comparisons have been made in the

context of land use change effects of N stocks, e.g. with afforestation programs and

conversion of forest to pasture landscapes. Reported soil N stocks ranged from 3.5 – 4.9

Mg ha-1 (0 – 15 cm) in a mountainous forest (Finzi et al., 1998), from 1.5 – 5.0 Mg ha-1

(0 – 30 cm) in an Amazon tropical forest (Neill et al., 1997) and 18.2 Mg ha-1 (0 – 100

cm) for a forest in southern Ethiopia (Demessie et al., 2011).

Although most studies considered only C, soil C and N cycles are

interdependent. Nitrogen is an essential nutrient for plants, and a constituent of the

Rubisco enzyme responsible for photosynthesis (Raines and Lloyd, 2001). Carbon

assimilated during photosynthesis is stored in plant tissues, transported to microbial

symbionts in the rhizosphere, respired or released into the soil (Prietzel and

Christophel, 2014). Nitrogen introduced to soil as plant litter, is a nutrient for

microorganisms, which decompose the organic matter and release C to the atmosphere

as CO2. Under increased CO2, net primary productivity (NPP) of many ecosystems is


Page | 43

expected to initially increase, causing N to be immobilized in biomass, depleting soil N,

and thereby increasing C:N ratios in soils and slowing rates of mineralization (Adams et

al., 2004). The feedback effect could ultimately limit responses to increased CO2. It is

imperative that soil N stocks and C:N ratios are assessed in order to determine the C

sequestration potential of soils (Finzi et al., 2006; Luo et al., 2006).

The south western foothills of the Bhutan Himalayas are characterized by a

range of forest types being present over short distances due to a steep altitudinal

gradient. Variation in altitude creates a gradient in abiotic factors such as temperature,

moisture and solar radiation, which in turn influence forest composition (Laughlin and

Abella, 2007) and soil organic carbon (SOC) (Jobbagy and Jackson, 2000; Singh et al.,

2011). Numerous studies have suggested that variation in abiotic factors as well as

species composition contribute to variations in C density in soil and biomass (Jobbagy

and Jackson, 2000; Lamlom and Savidge, 2003; Martin and Thomas, 2011; Zhu et al.,

2010). As a consequence of the steep altitudinal gradients in Bhutan, there are climatic

regimes at high altitude that are similar to those of widely separated latitudinal zones

(Beniston et al., 1997), which makes mountainous ecosystems more vulnerable to

climate change, the effects of which can be more rapid and severe than at lower

altitudes.

Most studies report that soil C stocks increase with altitude. Such results have

been reported for different regions and different forest types, including spruce, fir and

mixed hardwood forest in the USA (Garten Jr and Hanson, 2006), mixed broad-leaved

and pine forest in Nepal (Maraseni and Pandey, 2014) and mixed broad-leaved and pine

forest, grasslands, agricultural and horticultural lands in the western Himalayas in India

(Singh et al., 2011). In contrast, other researchers have reported decreasing soil C

stocks (Kumar et al., 2013) or no relationship of soil C stock (Godgift et al., 2014;

Tewksbury and Van Miegroet, 2007) with increasing altitude. We thus synthesized all

relevant published literature to investigate the global relationship of soil C with altitude.

In order to understand the influence of altitude and forest type on potential to sequester

or release C, and to inform forest management decisions, we aimed to:


Page | 44

i. Quantify the total soil C and N stocks at different depths along an altitudinal

gradient in Bhutan;

ii. Investigate the influence of forest composition, altitude and biomass input

on the soil C and N; and

iii. Synthesize a global relationship between soil C and altitude in different eco-

regions of the world via a meta-analysis of published literature.

3.2 Material and Methods

3.2.1 Study area

Our sampling sites are in the eastern foothills of the Himalayas, more

particularly in south western Bhutan (Fig. 3.1). Soil and plant samples were taken along

a transect running from the foothills at an elevation of 317 m a.s.l. (N 26° 51´, 89° 23´

E) to the mid-hills, where the elevation reached to 3300 m a.s.l. (N 26° 59´, 89° 32´ E).

A land cover map of Bhutan (2009) and Google Earth maps were used to geo-reference

the transect, as well as the sampling sites, prior to field studies. Based on vegetation

composition, forests were zoned into five types (Table 3.1), following the classification

proposed by (Ohsawa, 1987). i) tropical forest (TF) from 317 – 900 m a.s.l.), ii) sub-

tropical forest (STF) from 900 – 1870 m a.s.l., iii) warm temperate broadleaf forest

(WTBLF) from 1870 – 2450 m a.s.l., iv) cool temperate broadleaf forest (CTBLF) from

2450 – 3000 m a.s.l., and v) cold temperate forest (CTF) from 3000 – 3300 m a.s.l. The

study sites fall within two of the four Himalayan tectono-stratigraphic zones found in

Bhutan. TF and STF falls under the Lesser Himalayan formation composed of low

grade meta-sedimentary rocks, including quartzite, phyllite, and limestone (Long et al.,

2011; Tobgay et al., 2010). WTBLF, CTBLF and CTF fall under the Greater

Himalayan formation which consists of orthogneiss and meta-sedimentary rocks

(Gansser, 1983; Tobgay et al., 2010).


Page | 45

Fig. 3.1 Geographic location of the study area with main land use land cover types

Forested land has traditionally been used for grazing by migratory cattle herds.

In the last few decades, some forested areas have been clear felled in a number of

localities. Clear felled patches were avoided and not included in the study as they are

not representative of the natural C stocks for the forests. The southern foothills have

tropical climate with an average annual rainfall of 4600 mm and average annual

temperature of 22.9 °C (2009, Department of Hydro-Met Services, Bhutan). In the mid-

hills, between 2000 and 3000 m a.s.l., annual precipitation is about 3500 mm, and

summer temperatures can reach as high as 29 °C while the winter temperatures can

drop as low as 3 °C in December (Wangda et al., 2009).

Table 3.1 Forest characteristics along the altitudinal gradient

Altitude

(m a.s.l.)

Forest No. of Species H´ Density

(trees ha-1)

BA

(m2 ha-1)

MAT

(°C)

317 - 900 TF 33 2.96 313 17.8 22.9

900 - 1870 STF 54 3.47 383 30.1 15.4

1870 - 2450 WTBLF 42 3.21 433 41.1 13.3

2450 - 3000 CTBLF 47 3.14 824 46.3 10.9

3000 - 3300 CTF 10 1.96 511 66.6 5.5

TF = Tropical forest, STF = Sub-tropical forest, WTBLF = Warm temperate broadleaf forest,

CTBLF = Cool temperate broadleaf forest, CTF = Cold temperate forest, No. of species =


Page | 46

number of tree species surveyed for each forest type, H´ = Shannon diversity index, Density =

number of trees per ha, BA = basal area of trees (m2 ha-1), MAT = mean annual temperature

(°C).

3.2.2 Soil sampling and analysis

A total of 40 soil profiles, spaced at 75 m altitude interval, were dug along the

transect. Prior to digging the soil profile a 1 m × 1 m plot was laid out and all leaf litter

was collected, weighed and a subsample taken for further analysis. Soil samples were

collected from 0 to 100 cm depth from each profile. Soil profiles were sampled and

described as per guidelines of the Soil and Plant Analytical Laboratory (SPAL),

Simtokha Bhutan (SPAL, 1993).

Duplicate bulk density (BD) samples were taken from each of the four depth

categories i.e. 0 – 10 cm measured from the top of the soil after removal of the leaf

litter, 10 – 30 cm, 30 – 60 cm and 60 – 100 cm. Subsequently bulk soil samples of

about 1 kg were collected from each of the four depth categories. Soil samples were

oven dried at 40 °C, gently crushed by hand, weighed and passed through a 2 mm sieve.

The proportion of the > 2 mm fraction which consisted mostly of pebbles was

quantified by weighing and then discarded.

Soil pH and electrical conductivity (EC) were measured in a 1:5 soil:water

suspension, after shaking for 30 minutes (Rayment and Higginson, 1992). Particle-size

analysis was based on the pipette method, cation exchange capacity (CEC) was

measured with 1M ammonium acetate at pH 7. Bulk density was determined by drying

a known volume of soil to constant weight at 105 °C (SPAL, 1993). For total C and N

analyses, representative soil samples were finely ground (< 53 µm) using a Fritsch

Pulverisette 2 Mortar Grinder Mill (RETSCH GmbH, Haan, Germany). The total C and

N were analysed using a Vario MAX CNS analyzer (Elementar Analysensysteme

GmbH).

Soil C and N stocks (Mg C and N ha-1) in each layer were calculated

(Sanderman et al., 2011) as follows:


Page | 47

C content �Mg C ha-1� =

C %100

× BD �g cm-3� × layer thickness (m) × correction for units �1010 cm2

ha× Mg

106g� ×

correction for gravel content � g < 2 mm g total soil

� Equation 3.1

To compare our C and N stock data with other studies, a cubic spline function

in Microsoft Excel was used to interpolate C and N stocks data to match with soil

depths reported in other studies.

3.2.3 Forest inventory, biomass sampling and analysis

Along the transect, we developed a vegetation inventory by sampling 30 m ×

30 m plots at 150 m altitudinal intervals. In total, 20 vegetation inventory plots were

surveyed. All trees above 10 cm diameter at breast height (DBH) measured at 1.3 m

from the uphill side were identified and measured for diameter and height. Diameters

and heights were used to calculate bole volumes using equations developed for Bhutan

(Ellerbrock and Gerke, 2013). A global database (Zanne et al., 2009) was used as the

source of tree density data and to convert volumes to mass. Default values of biomass

expansion factors (IPCC, 2003) were used to calculate tree biomass. Within the 30 m ×

30 m inventory plots, three 1 m × 1 m sub-plots were randomly chosen to harvest the

herb layer and collect the leaf litter and dead wood (< 5 cm diameter).

Harvested plants were segregated into stems, branches and foliage and each

biomass category was weighed in the field. Subsamples of each biomass category were

weighed and taken to the laboratory and oven dried at 60 °C to constant weight. Dry

weights of subsamples were used to estimate dry weight of each biomass category.

Dried biomass fractions were coarsely ground to a powder by using Philips HL

1606/00 mixer-grinder and then a representative aliquot was finely ground using Retsch

MM400 Mixer Mill (RETSCH GmbH, Haan, Germany). The finely ground portion of

the sample was used for C and N analyses in a Thermo Finnigan Delta V isotope ratio

mass spectrometer coupled to ConfloIV and FlashHT peripherals (Thermo Fisher

Scientific, Bremen, Germany). Total C and N contents for each of the biomass


Page | 48

components of the different forest types were estimated as product of dry weight and C

and N concentrations.

3.2.4 Statistical analysis

IBM SPSS Statistics 21 was used to perform statistical analyses. Total soil C

and N concentrations and stocks at different depth categories for the different forest

types along the altitudinal gradient were compared using multivariate GLM

(generalized linear model) combined with a Tukey HSD post hoc test. Relationships

between total soil C and N stocks along the altitudinal gradient were assessed using a

linear regression analysis.

For biomass components, a similar multivariate GLM analysis was used to

compare C and N concentrations and stocks for different forest types along the

altitudinal gradient. Correlations (Pearson product-moment correlation coefficients)

among altitude, environmental, edaphic, allometric parameters and soil C and N

concentrations and stocks were considered significant at P < 0.05.

3.2.5 Meta-analysis for soil C along the altitudinal gradient

We compiled data from 28 studies (Supplement 3.1) that investigated SOC

changes with altitude in forest landscapes. These included studies originated from

Europe (n = 2), Africa (n = 3), the Americas (n = 7) and Asia (n = 16). Experimental

data included altitudinal ranges from sea level to 4800 m a.s.l., a MAT range from –7 °

to 26 °C and a MAP range from 30 to 4800 mm. The data provided 36, 15, 13, 9 and 3

observations for correlation between SOC (stocks or concentration at variable depths)

and altitude, MAT, MAP, BA and leaf litter, respectively. Meta-analysis was based on

correlations. When studies reported p-values instead of correlation coefficients,

p-values were converted to standard normal deviates Z and then transformed to

correlation coefficients using Meta Calc (Rosenberg et al., 1999). Effect size and

variance of each individual study was calculated using Meta Win v.2.1. based on

reported correlation coefficient (r) and sample sizes. Individual effect sizes and variance

were used to generate mean effect size and a bias-corrected 95% confidence interval

(CI) by bootstrapping (4999 iterations). Correlations were considered significant if the


Page | 49

95% of CI did not overlap with zero. The mean effect size and CI from Fisher’s z

metric were reconverted to correlation coefficients using Meta Calc for interpretation of

results.

3.3 Results

3.3.1 Influence of forest type and soil depth on carbon and nitrogen stocks in the

soil on the Bhutan transect.

Soil C and N stocks increased significantly with altitude. For soil depths 0 − 30

cm and 0 − 100 cm, altitude explained 55% and 73%, respectively, of the variation in

total C stocks. Correspondingly, for every 100 m rise in altitude C stocks increased by

4.3 Mg C ha-1 for soil depths of 0 − 30 cm and 12.4 Mg C ha-1 for 0 − 100 cm. The

intercepts for C stored for soils depths 0 − 30 cm was 30.025 Mg C ha-1 and for 0 − 100

cm was 49.639 Mg C ha-1 (Fig. 3.2 a&c). Nitrogen stocks increased with altitude

similar to C stocks. Altitude explained 35% of the variation in N stocks for soil depths 0

− 30 cm, and 47% for the depths 0 − 100 cm. For every 100 m rise in altitude, N stocks

increased by 0.23 Mg ha-1 for 0 − 30 cm and by 0.67 Mg ha-1 for 0 − 100 cm. The

intercepts for N stored for soils depths 0 − 30 cm was 4.037 Mg C ha-1 and for 0 − 100

cm was 8.699 Mg C ha-1 (Fig. 3.2 b&d).


Page | 50

Fig. 3.2 Relationship between total C and N stocks (Mg ha-1) with altitude on the

Bhutan transect for soil depths from 0 – 30 cm (a and b) and 0 – 100 cm (c and d). (a)

total carbon stock and (b) total N stock from 0 – 30 cm; (c) total carbon stock and (d)

total nitrogen stock from 0 – 100 cm.

Total soil C and N stocks varied significantly with forest type and soil depth

(Table 3.2). Total C stock for the top 30 cm soil in TF was 49 Mg ha-1 and in STF was

91 Mg ha-1, which were significantly less than stocks in WTBLF, CTBLF and CTF

(130 − 187 Mg ha-1). Of the total C stock contained in 0 − 100 cm depth, the 0 − 30 cm

interval constituted between 35 and 46% while the 0 − 60 cm interval provided 64 to

74%. Deeper soils (e.g. 30 – 100 cm and 60 – 100 cm) contributed between 24 to 65%

of the total C in 0 – 100 cm.


Page | 51

Table 3.2 Total carbon and nitrogen stocks (Mg ha-1) and percentages of total at various

depths in soils from different forest zones and altitudes in Bhutan Altitude (m a.s.l.)

Forest Zones

Total: 0 -100 cm 0 - 10 cm 0 - 30 cm 30 - 100 cm 60 - 100 cm

Carbon

(Mg ha-1) (Mg ha-1) % (Mg ha-1) % (Mg ha-1) % (Mg ha-1) % 317-900 TF 114.5 20.4 17.8 49.3 43.1 65.2 56.9 31.6 27.6 900-1870 STF 217.6 35.7 16.4 91.3 42.0 126.3 58.0 66.1 30.4 1870-2450 WTBLF 326.1 49.6 15.2 130.3 40.0 195.8 60.0 117.4 36.0

2450-3000 CTBLF 408.2 51.0 12.5 143.8 35.2 264.4 64.8 135.9 33.3

3000-3300 CTF 403.5 78.5 19.5 186.5 46.2 217.1 53.8 97.5 24.2 Nitrogen

(Mg ha-1) (Mg ha-1) % (Mg ha-1) % (Mg ha-1) % (Mg ha-1) % 317-900 TF 10.8 1.9 17.6 4.5 41.7 6.3 58.3 3.1 28.7 900-1870 STF 18.6 3.1 16.7 7.6 40.9 11.0 59.1 5.7 30.6 1870-2450 WTBLF 25.7 4.0 15.6 10.5 40.9 15.2 59.1 9.1 35.4 2450-3000 CTBLF 26.4 3.6 13.6 9.5 36.0 17.0 64.4 8.7 33.0 3000-3300 CTF 25.7 5.4 21.0 12.3 47.9 13.4 52.1 5.8 22.6

Similar to C stocks, total N stocks in the top 30 cm soil for TF (4.5 Mg ha-1)

and STF (7.6 Mg ha-1) were significantly lower than the other three higher altitude

forest types. Total N stocks in the 0 – 30 cm soil were 10.5 Mg ha-1 in the WTBLF, 9.5

Mg ha-1 in the CTBLF and 12.3 Mg ha-1 in the CTF. Proportions of N stored in each

soil depth category were similar to those of C stocks (Table 3.2).

C:N ratios of soil were highly correlated with altitude (r = 0.69, P = 0.01) and increased

from 9.8 for TF to 15.4 for CTF. Increases in C:N ratio with altitude were greater for

deeper soils compared to surface soils. Additionally, the C:N ratios for surface soils (0

– 10 cm) were more variable than for deeper soils (Fig. 3.3).


Page | 52

Fig. 3.3 Relationship between soil C:N ratio and altitude for different soil depth

categories

In the present study, proportionally more C was stored at shallower soil depth

when compared to forest soils from other studies (Fig. 3.4). The proportion of total C (0

− 100 cm) found in the 0 − 20 cm interval ranged from 31 to 69% for the present study

compared to 33 to 52% from other studies. Whereas the proportion of total C to 100 cm

that is in the deeper soils (60 – 100 cm) ranged from 3.5 to 36% for the present study,

which is more variable than 9.4 to 14.2% for other similar studies.

Fig. 3.4 Average carbon stocks in the present study and from other forest ecosystems

around the world (Debasish et al., 2012; Jobbagy and Jackson, 2000).


Page | 53

3.3.2 Influence of forest type and altitude on carbon and nitrogen concentrations in

understorey live biomass, canopy dead wood and leaf litter

Carbon concentrations in understory foliage and stem wood, canopy dead

wood and leaf litter varied significantly with forest types (Table 3.3) and altitude (Fig.

3.5). All biomass components from CTF had the greatest C concentrations (45.4 −

48.6%) and TF showed the least C concentrations (39.7 − 43.0%). Increasing elevation

resulted in increasing C concentrations in deadwood from 42% to 48.6%

(corresponding to an increase of 2.0% for every 1000 m) (Fig. 3.5a), in leaf litter from

42% to 47.8% (corresponding to an increase of 2.3% for every 1000 m) (Fig. 3.5b), in

understorey foliage from 42% to 45% (corresponding to an increase of 2.1% for every

1000 m) (Fig. 3.5c) and in stem wood from 43% to 46% (corresponding to an increase

of 1.0% for every 1000 m) (Fig. 3.5d) for the current study transect. Regression

analysis revealed that altitude accounted for changes in C concentrations of 45% in

deadwood, 27% in leaf litter, 38% in foliage, and 14% in stem wood. Additionally, total

leaf litter C density (g m-2) increased significantly from low to high altitude forest (TF =

62.0 g m2, STF =109.1 g m-2, WTBLF = 189.9 g m-2, CTBLF = 274.0 g m-2 and CTF =

166.2 g m-2; Table 3.3). There was no significant difference in the total C density with

altitude for other understorey biomass components.


Page | 54

Table 3.3 Nitrogen and carbon contents (Mean ± S.D.) on mass (%) and volume

(g m-2) basis in the understory biomass of different forest zones in Bhutan

Forest Zone C (%) N (%) C (g m-2) N (g m-2) Biomass (g m-2)

Foliage

TF 39.7a ± 3.3 2.9a ± 0.7 147.6a ± 109.1 9.6a ± 5.3 372.5a ± 273.2 STF 43.2b ± 2.2 2.8a ± 0.6 156.5a ± 69.2 10.6a ± 5.6 360.6a ± 154.5 WTBLF 43.2b ± 2.4 2.9a ± 0.5 166.9a ± 55.6 10.9a ± 3.4 388.3a ± 133.6 CTBLF 45.7b ± 1.3 2.6a ± 0.7 97.6 a ± 61.7 6.1a ± 5.7 166.7b ± 124.1 CTF 45.4b ± 0.2 1.7a ± 0.1 60.6a ± 19.2 2.3a ± 0.6 133.3b ± 41.6 Stem wood

TF 43.0a ± 2.9 0.8 a ± 0.3 127.7 a ± 138.8 1.8a ± 1.1 286.9a ± 303.5 STF 45.4b ± 2.0 0.8 a ± 0.2 76.3ab ± 38.5 1.5a ± 0.9 161.5ab ± 86.4 WTBLF 44.2ab ± 1.2 0.8 a ± 0.3 85.0ab ± 42.8 1.7a ± 1.1 190.8ab ± 94.1 CTBLF 45.8b ± 1.6 1.0 a ± 0.3 36.1b ± 26.4 0.7a ± 0.4 63.2b ± 57.9 CTF 46.1ab ± 0.7 0.7 a ± 0.2 44.2ab ± 25.0 0.8a ± 0.6 96.3ab ± 56.2 Deadwood

TF 42.0a ± 4.2 0.9ab ± 0.2 108.9 a ± 67.6 2.4a ±1.3 264.3a ± 172.9 STF 45.1b ± 2.5 1.1a ± 0.2 114.8a ± 77.1 2.9a ± 2.0 249.6a ± 177.4 WTBLF 46.1b ± 1.5 1.1a ± 0.1 174.8a ± 140.7 4.2a ± 3.4 378.4a ± 306.1 CTBLF 47.0b ± 1.0 1.0ab ± 0.2 117.2a ± 78.0 2.6a ± 2.3 249.6a ± 164.7 CTF 48.6b± 0.0 0.6b ± 0.0 109.1a ± 35.8 1.4a ± 0.4 224.2a ± 73.6

Leaf litter

TF 41.9a ± 4.8 1.2ac ± 0.2 62.0a ± 24.8 1.8a ± 0.8 148a ± 54.4 STF 44.9ab ± 3.7 1.7b ± 0.3 109.1a ± 56.8 4.3a ± 2.4 234.9a ± 123.5 WTBLF 44.8ab ± 2.8 1.8b ± 0.1 189.9b ± 96.9 7.8b ± 4.1 418.3b ± 205.2 CTBLF 46.4b ± 3.0 1.5ab ± 0.2 274.0b ± 111.3 9.2b ± 4.8 547.1b ± 222.9 CTF 47.8b ± 1.1 0.9c ± 0.2 166.2ab ± 28.4 3.3ab ± 1.3 347.6ab ± 60.9

Different letters within each column indicate significant difference between the forest types for

the measured parameters (P < 0.05).


Page | 55

Fig. 3.5 Carbon concentration in (a) dead wood (< 5 cm diameter), (b) leaf litter, (c)

understory foliage and (d) stem wood at different altitudes. Regression lines indicate

statistical significance at P = 0.05.

Nitrogen concentrations in branch wood (P = 0.473) and foliage (P = 0.087)

did not vary significantly between the forest types. However, N concentration differed

significantly among forest types for dead wood (P = 0.001) and leaf litter (P = < 0.001;

Table 3.3). In accordance with total C density, total N density in leaf litter also differed

significantly among TF, STF, WTBLF, CTBLF and CTF (density ranged from 1.8 g m-

2 to 9.2 g m-2). However, total N (g m-2) in the dead wood and live biomass components

did not differ among forest types.

Biomass contributions from deadwood less than 5 cm in diameter did not vary

among forest types. While the understorey foliage biomass for TF, STF and WTBLF

were significantly greater than the CTBLF and CTF. In contrast, the biomass from the


Page | 56

leaf litter for TF and STF was significantly less than the WTBLF and CTBLF (Table

3.3). For trees > 10 cm DBH, the biomass increased by 8.39 Mg C ha-1 for every 100 m

rise in altitude (Fig. 3.6).

Fig. 3.6 Relationship between tree (> 10 cm DBH) biomass with altitude

3.3.3 Influence of edaphic parameters on the carbon and nitrogen stocks in soil

Soil C and N stocks were significantly (P < 0.01) correlated with altitude (Fig.

3.2), leaf litter (g m-2), clay (%), CEC (not for N stock) and soil C:N ratios (Fig. 3.7). In

contrast, C and N stocks were negatively correlated with soil pH. However regression

analysis of soil pH with C and N stocks were not significant. Species richness of forests

were significantly correlated with C and N concentrations (P = 0.05), but not with C

and N stocks. Although BA of the forest was positively correlated with altitude, it was

neither significantly correlated with soil C and N stocks, nor with soil C and N

concentrations (Table 3.4).


Page | 57

Fig. 3.7 Relationship between C and N stock with leaf litter, clay percent, CEC, soil

C:N ratio and bulk density. Regression lines indicate statistical significance at (P =

0.05).


Page | 58

3.3.4 Meta-analysis of studies on soil carbon trend along altitudinal gradients

The present study and a majority of previous studies show positive correlations

of soil C with altitude. A few exceptions showed no relation or negative correlation

(Fig. 3.8a). Overall results of meta-analysis highlight a positive correlation of SOC with

altitude (r = 0.38, CI = 0.27 to 0.49, Fig. 3.8b). When altitude was binned into two

categories of high (> 1500 m a.s.l.) and low (< 1500 m a.s.l.), the correlation coefficient

did not differ significantly. SOC was negatively correlated with MAT (r = –0.49, CI = –

0.66 to –0.28) and positively correlated with MAP (r = 0.47, CI = 0.22 to 0.64). When

MAT and MAP were binned into high and low classes, meta-analysis suggested no

significant change in the strength of correlation between SOC and MAT or MAP.

While forest BA was not related to SOC, quantities of leaf litter (kg m-2) on the forest

floor were positively correlated with SOC (r = 0.35, CI = 0.01 to 0.63) (Fig. 3.8b), as

found in the present study.

Fig. 3.8 (a) Effect size (Zr) for Fisher’s z and 95% CI for SOC with altitude for

individual studies. The mean effect size is significantly different when its 95%

confidence interval does not bracket zero. (b) Forest plot showing the correlation

coefficient and 95% CI for total soil C and the parameters leaf litter quantity, basal area,

mean annual precipitation, mean annual temperature and altitude. The number of

observations per parameter is listed in brackets.


Page | 59

Table 3.4. Pearson correlation coefficients between soil (0 – 100 cm) and biomass properties

C

(Mg ha-1) N

(Mg ha-1) Alt Temp (°C )

Soil C (%)

Soil N (%) pH EC CEC

Clay (%)

Sand (%)

Silt (%)

Sp. richness

Basal area

Leaf litter C/N

C (Mg ha-1) 1 N (Mg ha-1) .72** 1 Alt (m) .57** .24* 1 Temp (°C ) -.50** -.17 -.91** 1 Soil C (%) .32** .17 .62** -.54** 1 Soil N (%) .25* .16 .48** -.44** .97** 1 Soil pH -.15 -.09 -.36** .44** -.49** -.46** 1 EC -.11 -.09 .20 -.13 .53** .55** -.18 1 CEC .35** .15 .62** -.52** .88** .83** -.39** .56** 1 Clay (%) .26* .28* -.04 .06 -.02 -.02 -.07 .05 .22* 1 Sand (%) -.13 -.12 .06 -.03 -.05 -.06 .07 -.19 -.30** -.64** 1 Silt (%) -.01 -.05 -.05 .<01 .08 .09 -.04 .21 .24* .14 -.85** 1 Sp. richness .18 .10 .33** -.22 .24* .26* .05 .20 .22* -.13 .12 -.07 1 Basal area (m2 ha-1) .20 .< 01 .69** -.71** .30** .20 -.24* .20 .35** -.13 -.02 .11 .22* 1 Leaf litter (g m-2)

.47** .20 .73** -.66** .56** .47** -.30** .19 .57** -.04 .07 -.05 .55** .37** 1 C/N ratio .53** .23* .77** -.68** .34** .15 -.31** -.06 .37** -.02 .02 -.01 .14 .54** .63** 1

* Correlation are significant at 0.05 and ** correlation are significant at 0.01 level.


Page | 60

3.4 Discussion

3.4.1 Soil carbon and nitrogen trends with altitude

Soil C and N stocks and concentrations were significantly correlated with

altitude, as well as temperature and edaphic parameters, which are in turn strongly

determined by altitude. Correlation coefficients of altitude with C and N stocks were

less than for C and N concentrations, mainly due to variation in BD and its effect on

calculated C and N stocks. BD decreased with increasing altitude. Weak correlations

of C and N stocks with clay content were not repeated for sand and silt contents.

Increases in soil C and N stocks with altitude were mainly influenced by increased C

and N input to soils from forest canopies, evident from the increase in basal area of

the forest with altitude and the effect of temperature, rather than by soil texture per

se. CEC was also highly correlated with both C and N concentrations and stocks, and

CEC were generally high in soils with high organic content (Parfitt et al., 1995),

suggesting that increases in C and N with altitude are due to increased organic matter

content. Increases with altitude in basal area, forest density (Table 3.1) and leaf litter

(Table 3.3) and tree biomass (Fig. 3.6) supports this conclusion.

Increasing soil C and N stocks with altitude (and with decreasing

temperature) is in agreement with meta-analysis wherein soil C increased with

altitude (r = 0.40, CI = 0.29 to 0.50). Dieleman et al. (2013) compared SOC stocks

for tropical forest along an altitudinal gradient for 10 studies and, similar to our

results, found a positive correlation (r = 0.43, P < 0.001). Altitude and temperature

showed very similar correlation coefficients, pointing to temperature being the main

driver for the C and N cycles and soil chemistry along the altitudinal gradient. In the

global data set, temperature was the most important environmental parameter after

precipitation (Fig. 3.8b), which was not available for the current study. Microbial

decomposition rates decrease as altitude increases and temperatures fall.

Accordingly, SOC is typically less in subtropical forest compared to temperate forest

(Yang et al., 2007). Besides the effect of temperature on microbial activity, total

microbial biomass, composition and activity are influenced by soil pH. For two

similar tropical forests in the Peruvian Amazon, the less acidic soil had a greater


Page | 61

diversity of soil bacteria (Fierer and Jackson, 2006). In our study (Table 3.5), slower

rates of organic matter decomposition at higher altitudes were accompanied by

increasing soil acidity, and increased stocks of C and N (see also (Garten Jr and

Hanson, 2006).

Soil C:N ratio also increased with altitude. The dominant influence of

temperature on soil respiration (Hopkins et al., 2012; Miller, 2000) ensures that at

lower elevations, where temperatures are greater, losses of C to the atmosphere are

maximized relative to those of N. At higher altitude with lower temperatures, organic

matter decomposition slows and more C is retained relative to N. In our study,

deeper soils (30 − 60 and 60 − 100 cm) at lower altitudes had lower C:N ratios than

the surface soils, suggesting a greater degree of decomposition and older SOM in

deeper soils. With increasing altitude the C:N ratios of deeper soils tend to be higher

than the surface soils (0 − 10 and 30 − 60 cm), probably due to decreased

decomposition − arguably, organic matter in deeper soils at higher elevations is a

result of a far reduced role of microbial turnover in the chemical nature of stored

organic matter. Additionally to microbial processes, DOC may have contributed to

the increase of C:N ratio with altitude and depth. SOC is a major source of dissolved

organic carbon (DOC) (Ying et al., 2013) and C:N ratios in DOC were greater than

in SOC (Neff et al., 2000). DOC can contribute as much as 25% of the total C in the

soil (Neff and Asner, 2001). Coniferous forests are characterized by a greater flux of

DOC into the soil than deciduous forests (Hope et al., 1994). As high altitude

CTBLF and CTF were dominated by conifers, the larger C:N ratios at higher

altitudes and in deeper soil horizons may have been influenced by DOC percolation,

especially relative to the slower rates of C input and incorporation into the soil from

biomass decomposition (Kalbitz and Kaiser, 2008).

Total C and N stocks of the primary forests in the southern foothills of the

Himalayas in Bhutan are amongst the highest reported in the world, especially for

mountainous regions. Even though C and N are preferentially accumulated at

shallower depths, deeper soils (60 – 100 cm) store substantial amounts of C and N

stocks (from 24 – 36% and 22 – 35%, respectively of the total C and N contents) and

need to be considered for an accurate estimation of the C and N stocks in forest soils.


Page | 62

A study in south-western Australia has reported C stocks in soils from 0 to 5 m depth

to vary between 47 and 77% of the total C stock to bed rock (Harper and Tibbett,

2013). They also found organic C even at depths of 38 m, which suggests that there

can be substantial underestimations of C and N stocks when deeper soils are ignored.

In contrast, the greater C and N stocks in shallower soils at high altitude were mostly

contained in organic matter due to slower decomposition rate (i.e. high CEC and C:N

ratios) which is a good indication of the vulnerability of these stocks to increasing

temperatures under climate change scenarios.

3.4.2 Biomass carbon concentration with altitude

Carbon concentration in understorey biomass components along the

altitudinal gradient varied from 39.7% in the foliage of the TF to 48.6% in the dead

wood of the CTF. Accurate estimation of C stocks and C sink potential are crucial

for C accounting, and different methodological approaches may over or

underestimate C stocks, e.g. estimating standing dead tree biomass (Woodall et al.,

2012) and tree allometry (Chave et al., 2005). Generally, biomass C concentration is

considered to be 50% of the plant dry mass (EPA, 2011; IPCC, 2003; Keith et al.,

2014); however, this simplified generalization is not a good approximation for all

species and eco-systems and can lead to an erroneous estimation of C stock in forest

ecosystems. In order to improve the modeling of C stock in Australian forests

(FullCAM, National Inventory Report 2010), the C concentrations of different

biomass components were taken into account (Gifford, 2000). Biomass C

concentration may vary by as much as 10% even within same tree species because of

age and tree components (Fu et al., 2013). Carbon concentration has also been shown

to vary between species as well as between plant organs (Fu et al., 2013; Lamlom

and Savidge, 2003; Martin and Thomas, 2011; Thomas and Malczewski, 2007). In

the present study, most live biomass samples contained less than 47% C, and for all

biomass samples, concentrations were < 50%. Carbon concentrations differed among

the various components of understorey plants and dead biomass as well as with

altitude (Fig. 3.5). Similarly, tropical species have been previously noted to have

reduced concentrations of C than temperate species (Thomas and Martin, 2012). The

lower C concentrations in tropical forest biomass could lead to an overestimation of


Page | 63

C stocks in the tropics compared to high altitude forests, if a general biomass C

concentration of 50% of the dry mass is used. Even small stems (< 10 cm) have a

significant impact on C accounting (Preece et al., 2012). More detailed mapping of C

concentrations in different forest components of various forest types will improve

global C stock estimation of terrestrial forest ecosystems. Furthermore studies

conducted in regions with low MAT (< 8 °C) have reported an increasing trend of

soil C with increasing temperature (Callesen et al., 2003; Liski and Westman, 1997).

This implies that there is a threshold temperature wherein with increasing altitude

and decreasing temperature the soil C stocks will start to decrease, and can be further

explored along targeted altitudinal gradients.


Page | 64

References

Adams, M., Ineson, P., Binkley, D., Cadisch, G., Tokuchi, N., Scholes, M., Hicks,

K., 2004. Soil functional responses to excess nitrogen inputs at global scale.

AMBIO: A Journal of the Human Environment 33(8), 530-536.

Alexander, E.B., Mallory, J.I., Colwell, W.L., 1993. Soil-elevation relationships on a

volcanic plateau in the Southern Cascade Range, northern California, USA.

Catena 20(1–2), 113-128.

Ali, S., Hayat, R., Begum, F., Hussain, A., Hasan, N., Hameed, A., 2014. Altitudinal

distribution of soil organic carbon stock and its relation to aspect and

vegetation in the mountainous forest of Bagrot Valley, North Karakoram,

Gilgit-Baltistan. Journal of Biodiversity and Environmental Sciences 5(1),

199-213.

Beniston, M., Diaz, H.F., Bradley, R.S., 1997. Climatic change at high elevation

sites: An overview. Climatic Change 36(3-4), 233-251.

Callesen, I., Liski, J., Raulund-Rasmussen, K., Olsson, M.T., Tau-Strand, L.,

Vesterdal, L., Westman, C.J., 2003. Soil carbon stores in Nordic well-drained

forest soils—relationships with climate and texture class. Global Change

Biology 9(3), 358-370.

Campos C, A., Aguilar S, G., Landgrave, R., 2014. Soil organic carbon stocks in

Veracruz State (Mexico) estimated using the 1:250,000 soil database of

INEGI: Biophysical contributions. Journal of Soils and Sediments 14(5), 860-

871.

Charan, G., Bharti, V., Jadhav, S., Kumar, S., Acharya, S., Kumar, P., Gogoi, D.,

Srivastava, R., 2013. Altitudinal variations in soil physico-chemical

properties at cold desert high altitude. Journal of Soil Science and Plant

Nutrition 13(2), 267-277.




87-99.

Chhabra, A., Palria, S., Dadhwal, V.K., 2003. Soil organic carbon pool in Indian

forests. Forest Ecology and Management 173(1), 187-199.


Page | 65

Chuai, X.-W., Huang, X.-J., Wang, W.-J., Zhang, M., Lai, L., Liao, Q.-L., 2012.

Spatial variability of soil organic carbon and related factors in Jiangsu

Province, China. Pedosphere 22(3), 404-414.

Dai, W., Huang, Y., 2006. Relation of soil organic matter concentration to climate

and altitude in zonal soils of China. Catena 65(1), 87-94.

Debasish, S., Kukal, S.S., Bawa, S.S., 2012. Soil organic carbon stock and fractions

in relation to land use and soil depth in the degraded Shiwaliks hills of lower

Himalayas. Land Degradation & Development, 25/407-416.

Demessie, A., Singh, B.R., Lal, R., 2011. Soil carbon and nitrogen stocks under

plantations in Gambo District, Southern Ethiopia. Journal of Sustainable

Forestry 30(6), 496-517.

Dieleman, W.I.J., Venter, M., Ramachandra, A., Krockenberger, A.K., Bird, M.I.,

2013. Soil carbon stocks vary predictably with altitude in tropical forests:

Implications for soil carbon storage. Geoderma 204–205, 59-67.

Du B, Kang H, Pumpanen J et al. (2014) Soil organic carbon stock and chemical

composition along an altitude gradient in the Lushan Mountain, subtropical

China. Ecological Research, 29, 433-439.

Ellerbrock, R.H., Gerke, H.H., 2013. Characterization of organic matter composition

of soil and flow path surfaces based on physicochemical principles—A

review. Advances in Agronomy 121, 117-177.

EPA, A., 2011. Inventory of US greenhouse gas emissions and sinks: 1990-2009,

EPA 430-R-11-005.

Fierer, N., Jackson, R.B., 2006. The diversity and biogeography of soil bacterial

communities. Proceedings of the National Academy of Sciences of the United

States of America 103(3), 626-631.

Finzi, A.C., Moore, D.J.P., DeLucia, E.H., Lichter, J., Hofmockel, K.S., Jackson,

R.B., Kim, H.-S., Matamala, R., McCarthy, H.R., Oren, R., Pippen, J.S.,

Schlesinger, W.H., 2006. Progressive nitrogen limitation of ecosystem


Page | 66

processes under elevated CO2 in a warm temperate forest. Ecology 87(1), 15-

25.

Finzi, A.C., Van Breemen, N., Canham, C.D., 1998. Canopy tree - Soil interactions

within temperate forests: Species effects on soil carbon and nitrogen.

Ecological Applications 8(2), 440-446.

Fu, Y., Wang, X., Sun, Y., 2013. Carbon concentration variability of Larix olgensis

in North-Eastern China. Advance Journal of Food Science and Technology

5(5), 627-632.

Galka, B., Labaz, B., Bogacz, A., Bojko, O., Kabala, C., 2014. Conversion of

Norway spruce forests will reduce organic carbon pools in the mountain soils

of SW Poland. Geoderma 213, 287-295.

Gansser, A., 1983. Geology of the Bhutan Himalaya. Denkschr Schweiz Naturforsch

Ges.

Garten, C.T., Jr., Post, W.M., III, Hanson, P.J., Cooper, L.W., 1999. Forest soil

carbon inventories and dynamics along an elevation gradient in the southern

Appalachian Mountains. Biogeochemistry 45(2), 115-145.



Gifford, R.M., 2000. Carbon contents of above-ground tissues of forest and

woodland trees. National Carbon Accounting System technical report: no. 22.

Australian Greenhouse Office, Canberra.

Godgift, S., Henry, J.N., Pantaleo, K.T., Munishi, Deo, D.S., 2014. Carbon stocks of

Hanang forest, Tanzania: An implication for climate mitigation. Journal of

Ecology and Natural Environment 6(3), 90-98.

Harper, R.J., Tibbett, M., 2013. The hidden organic carbon in deep mineral soils.

PLANT AND SOIL 368(1-2), 641-648.

Hope, D., Billett, M.F., Cresser, M.S., 1994. A review of the export of carbon in

river water: Fluxes and processes. Environmental Pollution 84(3), 301-324.

Hopkins FM, Torn MS, Trumbore SE (2012) Warming accelerates decomposition of

decades-old carbon in forest soils. Proceedings of the National Academy of

Sciences. USA 109, E1753-E1761.


Page | 67

IPCC, 2003. Good practice guidance for land use, land-use change and forestry.,

Institute for Global Environmental Strategies (IGES), Hayama.



436.

Kalbitz, K., Kaiser, K., 2008. Contribution of dissolved organic matter to carbon

storage in forest mineral soils. Journal of Plant Nutrition and Soil Science,

Zeitschrift fur pflanzenernahrung und bodenkunde 171(1), 52-60.

Keith, H., Lindenmayer, D.B., Mackey, B.G., Blair, D., Carter, L., McBurney, L.,

Okada, S., Konishi-Nagano, T., 2014. Accounting for biomass carbon stock

change due to wildfire in temperate forest landscapes in Australia. PloS one

9(9), e107126.

Kumar, S., Kumar, M., Sheikh, M., 2013. Carbon stock variation of Pinus roxburghii

Sarg. Forest along altitudes of Garhwal Himalaya, India. Russian Journal of

Ecology 44(2), 131-136.

Lamlom, S.H., Savidge, R.A., 2003. A reassessment of carbon content in wood:

variation within and between 41 North American species. Biomass and

Bioenergy 25(4), 381-388.

Laughlin, D.C., Abella, S.R., 2007. Abiotic and biotic factors explain independent

gradients of plant community composition in ponderosa pine forests.

Ecological Modelling 205(1–2), 231-240.

Lemenih, M., Itanna, F., 2004. Soil carbon stocks and turnovers in various vegetation

types and arable lands along an elevation gradient in southern Ethiopia.

Geoderma 123(1–2), 177-188.

Li, P., Wang, Q., Endo, T., Zhao, X., Kakubari, Y., 2010. Soil organic carbon stock

is closely related to aboveground vegetation properties in cold-temperate

mountainous forests. Geoderma 154(3–4), 407-415.

Li, B., Tang, H., Wu, L., Li, Q., Zhou, C., 2012. Relationships between the soil

organic carbon density of surface soils and the influencing factors in differing

land uses in Inner Mongolia. Environmental Earth Sciences 65, 195-202.


Page | 68

Liski, J., Westman, J., 1997. Carbon storage in forest soil of Finland. 1. Effect of

thermoclimate. Biogeochemistry 36(3), 239-260.

Liu, W., Chen, S., Qin, X., Baumann, F., Scholten, T., Zhou, Z., Sun, W., Zhang, T.,

Ren, J., Qin, D., 2012. Storage, patterns, and control of soil organic carbon

and nitrogen in the northeastern margin of the Qinghai–Tibetan Plateau.

Environmental Research Letters 7(3), 035401.

Long, S., McQuarrie, N., Tobgay, T., Rose, C., Gehrels, G., Grujic, D., 2011.

Tectonostratigraphy of the Lesser Himalaya of Bhutan: Implications for the

along-strike stratigraphic continuity of the northern Indian margin.

Geological Society of America Bulletin 123(7-8), 1406-1426.

Longbottom, T.L., Townsend-Small, A., Owen, L.A., Murari, M.K., 2014. Climatic

and topographic controls on soil organic matter storage and dynamics in the

Indian Himalaya: Potential carbon cycle–climate change feedbacks. Catena

119(0), 125-135.

Luo, Y., Field, C.B., Jackson, R.B., 2006. Does nitrogen constrain carbon cycling, or

does carbon input stimulate nitrogen cycling? Ecology 87(1), 3-4.

Maraseni, T.N., Pandey, S.S., 2014. Can vegetation types work as an indicator of soil

organic carbon? An insight from native vegetations in Nepal. Ecological

Indicators 46(0), 315-322.

Martin, A.R., Thomas, S.C., 2011. A reassessment of carbon content in tropical trees.

PloS one 6(8), e23533.

Miller, C., 2000. Understanding the carbon/nitrogen ratio: Plants require a balance of

both, Countryside and Small Stock Journal. Countryside Publications, Ltd.,

Waterloo, pp. 65-66.

Neff, J.C., Asner, G.P., 2001. Dissolved organic carbon in terrestrial ecosystems:

Synthesis and a model. Ecosystems 4(1), 29-48.

Neff, J.C., Hobbie, S.E., Vitousek, P.M., 2000. Nutrient and mineralogical control on

dissolved organic C, N and P fluxes and stoichiometry in Hawaiian soils.


Neill, C., Melillo, J.M., Steudler, P.A., Cerri, C.C., Moraes, J.F.L.d., Piccolo, M.C.,

Brito, M., 1997. Soil Carbon and Nitrogen Stocks Following Forest Clearing


Page | 69

for Pasture in the Southwestern Brazilian Amazon. Ecological Applications

7(4), 1216-1225.

Ordóñez, J.A.B., de Jong, B.H.J., García-Oliva, F., Aviña, F.L., Pérez, J.V.,

Guerrero, G., Martínez, R., Masera, O., 2008. Carbon content in vegetation,

litter, and soil under 10 different land-use and land-cover classes in the

Central Highlands of Michoacan, Mexico. Forest Ecology and Management

255(7), 2074-2084.

Parfitt, R.L., Giltrap, D.J., Whitton, J.S., 1995. Contribution of organic matter and

clay minerals to the cation exchange capacity of soils. Communications in

Soil Science and Plant Analysis 26(9-10), 1343-1355.

Preece, N.D., Crowley, G.M., Lawes, M.J., Van Oosterzee, P., 2012. Comparing

above-ground biomass among forest types in the Wet Tropics: Small stems

and plantation types matter in carbon accounting. Forest Ecology and

Management 264, 228-237.

Prietzel, J., Christophel, D., 2014. Organic carbon stocks in forest soils of the

German Alps. Geoderma 221–222(0), 28-39.

Raich, J.W., Russell, A.E., Vitousek, P.M., 1997. Primary productivity and

ecosystem development along an elevational gradient on Mauna Loa,

Hawai'i. Ecology 78(3), 707-721.

Raines, C.A., Lloyd, J.C., 2001. C3 carbon reduction cycle. eLS.

Rayment, G., Higginson, F.R., 1992. Australian Laboratory Handbook of Soil and

Water chemical methods. Inkata Press Pty Ltd, Melbourne, Australia.

Razakamanarivo, R.H., Grinand, C., Razafindrakoto, M.A., Bernoux, M., Albrecht,

A., 2011. Mapping organic carbon stocks in eucalyptus plantations of the

central highlands of Madagascar: A multiple regression approach. Geoderma

162(3–4), 335-346.

Rosenberg, M.S., Adams, D.C., Gurevitch, J., 1999. MetaWin: Statistical software

for Meta-analysis (Sinauer Associates, Sunderland, MA, USA).

Sanderman, J., Baldock, J., Hawke, B., Macdonald, L., Massis-Puccini A, S, S.,

2011. National soil carbon research programme: Field and laboratory

methodologies. CSIRO, Canberra.


Page | 70

Shelukindo, H.B., Semu, E., Msanya, B.M., Singh, B.R., Munish, P.K.T., 2014. Soil

organic carbon stocks in the dominant soils of the Miombo woodland

ecosystem of Kitonga Forest Reserve, Iringa Tanzania. International Journal

of Agricultural Policy and Research 2(4), 167-177.

Singh, S.K., Pandey, C.B., Sidhu, G.S., Sarkar, D., Sagar, R., 2011. Concentration

and stock of carbon in the soils affected by land uses and climates in the

western Himalaya, India. Catena 87(1), 78-89.

SPAL, 1993. Soil and plant analytical laboratory: Soil analysis. Ministry of

Agriculture, Thimphu, Bhutan.

Tewksbury, C., Van Miegroet, H., 2007. Soil organic carbon dynamics along a

climatic gradient in a southern Appalachian spruce-fir forest. Canadian


Thomas, S.C., Malczewski, G., 2007. Wood carbon content of tree species in Eastern

China: Interspecific variability and the importance of the volatile fraction.

Journal of Environmental Management 85(3), 659-662.


Forests 3(2), 332-352.

Tobgay, T., Long, S., McQuarrie, N., Ducea, M.N., Gehrels, G., 2010. Using isotopic

and chronologic data to fingerprint strata: Challenges and benefits of variable

sources to tectonic interpretations, the Paro Formation, Bhutan Himalaya.

Tectonics 29(6).

Wangda, P., Norbu, L., Tashi, S., Gyeltshen, D., 2009. Change in forest structure and

diversity after the humane disturbances in the cool montane evergreen

broadleaved forest. Journal of Renewable Natural Resources of Bhutan 5(1),

55-67.

Woodall, C.W., Domke, G.M., Macfarlane, D.W., Oswalt, C.M., 2012. Comparing

field-and model-based standing dead tree carbon stock estimates across

forests of the US. Forestry 85(1), 125-133.

Yang, L., Pan, J., Shao, Y., Chen, J.M., Ju, W.M., Shi, X., Yuan, S., 2007. Soil

organic carbon decomposition and carbon pools in temperate and sub-tropical

forests in China. Journal of Environmental Management 85(3), 690-695.


Page | 71

Ying, C., Shaoqiang, W., Jingyuan, W., Peichl, M., Falahatar, S., 2013. Dissolved

organic carbon dynamics and controls of planted slash pine forest soil in

subtropical region in Southern China. Journal of Resources and Ecology 4(2),

105-114.

Zanne, A.E., Lopez-Gonzalez, G., Coomes, D.A., Ilic, J., Jansen, S., Lewis, S.L.,

Miller, R.B., Swenson, N.G., Wiemann, M.C., Chave, J., 2009. Data from:

Towards a worldwide wood economics spectrum. Dryad Data Repository.

Zhang, Y., Duan, B., Xian, J., Korpelainen, H., Li, C., 2011. Links between plant

diversity, carbon stocks and environmental factors along a successional

gradient in a subalpine coniferous forest in Southwest China. Forest Ecology

and Management 262(3), 361-369.

Zhu, B., Wang, X., Fang, J., Piao, S., Shen, H., Zhao, S., Peng, C., 2010. Altitudinal

changes in carbon storage of temperate forests on Mt Changbai, Northeast

China. Journal of Plant Research 123(4), 439-452.

Zimmermann, M., Meir, P., Silman, M., Fedders, A., Gibbon, A., Malhi, Y., Urrego,

D., Bush, M., Feeley, K., Garcia, K., Dargie, G., Farfan, W., Goetz, B.,

Johnson, W., Kline, K., Modi, A., Rurau, N.Q., Staudt, B., Zamora, F., 2010.

No Differences in Soil Carbon Stocks Across the Tree Line in the Peruvian

Andes. Ecosystems 13(1), 62-74.


Page | 72

Supplementary Information

Supplement 3.1 Data sources used in the meta-analysis that provided correlation

data on SOC with altitude, MAT, MAP, BA and forest floor leaf litter. Grouping Number of

studies Authors

Altitude 28 (Alexander et al., 1993; Ali et al., 2014; Campos C et al., 2014; Charan et al., 2013; Chuai et al., 2012; Dai and Huang, 2006; Dieleman et al., 2013; Du et al., 2014; Garten Jr and Hanson, 2006; Godgift et al., 2014; Kumar et al., 2013; Lemenih and Itanna, 2004; Li et al., 2012; Li et al., 2010; Liu et al., 2012; Longbottom et al., 2014; Maraseni and Pandey, 2014; Prietzel and Christophel, 2014; Raich et al., 1997; Razakamanarivo et al., 2011; Shelukindo et al., 2014; Singh et al., 2011; Tewksbury and Van Miegroet, 2007; Zhang et al., 2011; Zhu et al., 2010; Zimmermann et al., 2010)

MAT 10 (Campos C et al., 2014; Dai and Huang, 2006; Du et al., 2014; Garten et al., 1999; Lemenih and Itanna, 2004; Li et al., 2010; Liu et al., 2012; Prietzel and Christophel, 2014; Singh et al., 2011; Zhang et al., 2011)

MAP 8 (Campos C et al., 2014; Dai and Huang, 2006; Du et al., 2014; Lemenih and Itanna, 2004; Liu et al., 2012; Longbottom et al., 2014; Prietzel and Christophel, 2014; Singh et al., 2011)

BA 7 (Garten Jr and Hanson, 2006; Li et al., 2010; Raich et al., 1997; Tewksbury and Van Miegroet, 2007; Zhang et al., 2011)

Leaf litter 3 (Galka et al., 2014; Li et al., 2010; Zhang et al., 2011)


Page | 73

Supplement 3.2 Soil properties at different soil depths (Mean ± S.D) for forest types

in Bhutan. Forest Zone 0 - 10 cm 10 - 30 cm 30 - 60 cm 30 - 100 cm

Soil pH (1:5 H2O) TF 5.46 a ± 0.7 5.43 a ± 0.7 5.58 a ± 0.7 5.83 a ± 0.7 STF 4.85 b ± 0.4 4.96 b ± 0.4 5.15 b ± 0.4 5.34 b ± 0.5 WTBLF 4.71 b ± 0.2 4.92 b ± 0.2 4.97 b ± 0.2 5.24 b ± 0.3 CTBLF 4.66 b ± 0.3 4.74 b ± 0.2 4.91 b ± 0.1 5.09 b ± 0.1 CTF 4.88 ab ± 0.4 4.81 ab ± 0.2 4.88 b ± 0.1 5.13 ab ± 0.1

Electrical conductivity (1:5; µS cm- 1) TF 42 a ± 21 28 a ± 12 25 a ± 14 27 a ± 19.0 STF 52 a ± 26 36 a ± 16 35 a ± 27 21 a ± 6.0 WTBLF 64 a ± 30 31 a ± 11 23 a ± 5.0 20 a ± 7.0 CTBLF 64 a ± 39 60 b ± 32 43 a ± 29 29 a ± 8.0 CTF 67 a ± 15 54 ab ± 5.0 36 a ± 18 16 a ± 0.1

Sand (%) TF 49.4 a ± 7.2 51.4 a ± 8.4 51.8 a ± 9.7 53.6 ab ± 11.8 STF 39.3 a ± 12.8 36.5 b ± 11.0 40.4 a ± 9.3 43.1 a ± 11.9 WTBLF 51.6 a ± 17.1 53.1 a ± 16.5 52.8 a ± 21.8 57.8 b ± 18.8 CTBLF 46.9 a ± 18.7 41.8 ab ± 15.9 42.3 a ± 15.7 43.5 ab ± 17.8 CTF 74.3* 72.2* 80.4* 87.3*

Silt (%) TF 30.9 ab± 6.7 29.0 ac ± 6.1 27.5 ac ± 5.9 24.3 a ± 5.7 STF 39.3 a ± 8.1 40.9 bc ± 6.7 37.6 bc ± 8.5 38.3 b ± 12.0 WTBLF 28.8 b ± 8.2 24.7 a ± 4.7 24.2 a ± 7.1 22.4 a ± 5.1 CTBLF 34.8 ab ± 12.1 36.7 c ± 16.8 33.4 c ± 8.6 29.4 a ± 7.6 CTF 12.1* 11.8* 7.8* 3.1*

Clay (%) TF 19.6 a ± 5.5 19.5 a ± 5.5 20.6 a ± 7.1 22.0 a ± 8.3 STF 21.2 a ± 6.7 22.5 a ± 8.9 21.9 a ± 8.9 18.5 a ± 6.4 WTBLF 19.4 a ± 9.7 22.0 a ± 11.9 22.8 a ± 15.0 19.7 a ± 13.9 CTBLF 18.2 a ± 9.7 21.4 a ± 9.8 24.6 a ± 10.6 26.9 a ± 11.2 CTF 12.8* 15.8* 11.7* 9.5*

Cation exchange capacity (cmolc kg -1) TF 11.13a ± 7.0 8.69 a ± 4.5 6.7 a ± 2.3 8.53 a ± 5.4 STF 18.87 ab ± 9.6 14.98 a ± 7.2 11.6 ab ± 6.7 8.73 a ± 6.4 WTBLF 23.81 bc ± 7.8 17.52 ac ± 9.5 16.4 b ± 7.4 12.89 a ± 6.5 CTBLF 31.33 c ± 7.4 30.55 b ± 8.1 26.6 c ± 4.8 24.45b ± 6.1 CTF 33.23 bc ± 2.8 34.16 bc ± 9.6 21.1 bc ± 7.8 15.13 ab ± 2.4

Bulk density (g cm -3) TF 1.32 a ± 0.2 1.32 a ± 0.2 1.42 a ± 0.2 1.47 a ± 0.3 STF 0.91 b ± 0.2 0.98 b ± 0.2 1.08 b ± 0.2 1.07 b ± 0.2 WTBLF 0.70 c ± 0.1 0.89 b ± 0.2 0.92 b ± 0.2 0.98 b ± 0.2 CTBLF 0.56 c ± 0.2 0.62 c ± 0.2 0.68 c ± 0.1 0.72 c ± 0.1 CTF 0.52 c ± 0.1 0.52 c ± 0.1 0.82 bc ± 0.2 0.86 bc ± 0.1

* Standard deviation were not reported as data did not have any replicates.


Page | 74

Supplement 3.3 Total carbon and nitrogen stocks in Mg ha-1 (Mean ± S.D) at

various depths in soils from different forest zones in Bhutan. Altitude Forest Zones 0 - 10 cm 10 - 30 cm 30 - 60 cm 60 - 100 cm

(m a.s.l.) Total C (Mg ha -1)

317-900 TF 20.4 a ± 11.0 28.9 a ± 16.5 33.5a ± 19.2 31.6 a ± 18.6

900-1870 STF 35.7ab ± 12.6 55.5 ab ± 27.8 60.1ac ± 30.6 66.1a ± 47.4

1870-2450 WTBLF 49.6 b ± 13.0 80.6 bc ± 22.7 78.4ac ± 21.6 117.4 b ± 44.0

2450-3000 CTBLF 51.0 bc ± 5.8 92.7 c ± 24.2 128.5b ± 42.2 135.9 b ± 32.2

3000-3300 CTF 78.5 c ± 46.3 107.9 bc ± 59.9 119.5bc ± 58.2 97.5 ab ± 23.9

Total N (Mg ha -1 )

317-900 TF 1.9 a ± 0.9 2.5 a ± 1.2 3.1a ± 1.5 3.1a ± 1.8

900-1870 STF 3.1b ± 1.0 4.5 b ± 2.5 5.3 a ± 3.1 5.7 a ± 4.3

1870-2450 WTBLF 4.3 cd ± 1.2 6.2 b ± 1.8 6.1ab ± 1.8 9.1b ± 3.9

2450-3000 CTBLF 3.6 bc ± 0.8 5.8 b ± 1.4 8.3b ± 2.1 8.7b ± 2.7

3000-3300 CTF 5.4d ± 3.1 6.8 b ± 3.4 7.6 ab ± 4.1 5.8 ab ± 1.2

Different letters within each column indicate significant difference (P < 0.05) between the

forest types.

Chapter 4. Carbon and nitrogen isotopes in soil and vegetation

Page | 75

Chapter 4.

Elevation and light drive abundances of carbon and

nitrogen isotopes in soil and vegetation in the Himalayas2.

Abstract

Globally, soil carbon (C) content varies with climate, vegetation type and

productivity as well as soil properties. In turn, forest productivity varies with nitrogen (N)

availability, precipitation, temperature and altitude. While a previous study showed that

soil C and N stocks in the Eastern Himalayas are amongst the highest recorded, we have

limited knowledge of the driving influences. We used stable isotopes in combination with

the elemental composition of biomass and soil to help elucidate processes controlling C

and N input, turnover and stability in the soil (0 to 100 cm depth) along a transect from

317 to 3300 m a.s.l.. Forest overstorey biomass contributed significantly to soil C (δ13C in

overstorey biomass components were similar to soil δ13C). Changes in δ13C and δ15N with

soil depth (to 60 cm) were least at the highest altitude, cold temperate forest, suggesting

slow turnover of C and N at high altitudes and low temperatures, supported by increasing

soil C:N ratio and CEC. With depth, soil δ13C reflected decomposition, while soil δ15N

reflected decomposition, longer residence times and/or leaching and volatilization. Soil C

and N stocks in high elevation forests are vulnerable to losses via decomposition if long-

term temperatures continue to rise.

2 This chapter has been submitted to Ecosystems under the title “Elevation and light drive abundances of carbon and nitrogen isotopes in soil and vegetation in the Himalayas” in August 2016. Authors are Sonam Tashi, Claudia Keitel, Balwant Singh and Mark Adams


Page | 76

4.1 Introduction

Forest ecosystem composition and productivity are strongly influenced by

temperature (Moles et al., 2014; Reich et al., 2014) and nitrogen (N) availability (Curtis

et al., 2002; Fisichelli et al., 2015). In turn, forest production affects carbon (C) inputs to

soils (Garten Jr, 2011). Temperature not only alters the soil C dynamics via input from

biomass growth, but also influences microbial populations and organic C decomposition

processes within the soil (Mueller et al., 2015).

In mountainous ecosystems, the steep terrain leads to change in temperature at

land surface over relatively short distances and can be expected to induce variations in

soil C and N dynamics with altitude. Some studies suggest that soil C and N stocks

increase with altitude, and that temperature and precipitation are important drivers thereof

(Dieleman et al., 2013; Du et al., 2014; Tashi et al., 2016). Similarly, numerous studies

have reported increases in soil δ13C value with altitude (Bird et al., 1994; Marshall and

Zhang, 1994; Zhou et al., 2013; Zimmermann et al., 2012). δ13C value of whole soil often

reflects the δ13C composition of plant communities (Bai et al., 2012), and variation with

altitude in δ13C of biomass and soil provides insights to soil C dynamics (Bird et al.,

1996). Plant δ13C, for example, can reflect changes in several environmental factors that

change with altitude, such as temperature, soil moisture content, atmospheric pressure

and solar radiation (Beerling et al., 1996; Garten, 2006; Körner et al., 1991). In addition,

plant δ13C reflects physiological adjustments to environmental conditions, such as

drought (Farquhar and Richards, 1984; Marshall and Zhang, 1994). While a majority of

studies show positive relations between δ13C and altitude, some authors have reported

non-linear changes (Luo et al., 2006; Zhao et al., 2008), and a negative trend was

reported for mountainous terrain in China (Wang et al., 2010).

In contrast to a reasonably robust theoretical framework for δ13C (Farquhar et

al., 1982), for δ15N variations in plants and soil, we lack unifying theory. Several studies

have shown δ15N in soils and plants declines with increasing rainfall (Austin and Sala,


Page | 77

1999; Handley et al., 1999) and altitude (Liu and Wang, 2010). Similarly, due to its effect

on soil moisture, topography influences the δ15N of both plants and soil (Sutherland et al.,

1993). δ15N has also been proposed to reflect differences in the ‘openness’ of the N cycle;

Martinelli et al. (1999) evaluated δ15N in soils and plants from tropical and temperate

forests across various continents and attributed higher δ15N values in tropical forests to a

more open N system (i.e. more N outputs and inputs) than in temperate forest. More

generally, decomposition and mineralisation of organic matter result in lighter C and N

isotopes being preferentially lost, which leads to increases in δ13C and δ15N with soil

depth (Nadelhoffer and Fry, 1988). Such isotopic fractionations become greater as

temperature increases and are widely inferred as driving trends in soil δ13C and δ15N with

altitude (Schaub and Alewell, 2009).

In the Himalayas, stable carbon isotopes have been used to understand the

chemical weathering and erosion of dissolved organic C (Singh et al., 2005), to

reconstruct quaternary vegetation patterns (France-Lanord and Derry, 1994) and to

estimate the marine burial of organic C resulting from the Himalayan uplift (France-

Lanord and Derry, 1997). However, there is little information available on interactions

between soil and biomass δ13C and δ15N, and their interactions with altitude.

We studied mountainous terrain in the Bhutan Himalayas, which is characterised

by a highly variable climate (e.g. precipitation decreases from south to north (Burbank et

al. 2012) and diverse vegetation. In this study, we aimed to (i) identify factors that

determine patterns in δ13C and δ15N in biomass and soil with altitude, (ii) and identify the

drivers of C and N input, turnover and stability in the soil.

4.2 Material and Methods

4.2.1 Site description

The study area is located in the foothills of the eastern Himalayas, more

specifically in southern Bhutan. We chose a transect from the foothills of Phuentsholing

at an elevation of 317 m (N 26° 51´, 89° 23´ E) to Gedu top where the elevation reached


Page | 78

3300 m (N 26° 59´, 89° 32´ E). This transect covered tropical forest (TF) 317 – 900

m.a.s.l., subtropical forest (STF) 900 – 1870 m a.s.l., warm temperate broadleaf forest

(WTBLF) 1870 – 2450 m a.s.l., cool temperate broadleaf forest (CTBLF) 2450 – 3000 m

a.s.l., and cold temperate forest (CTF) > 3000 m a.s.l.. The forest zones are based on the

classification by Ohsawa (1987).

The lower altitudes (205 m) of the southern foothills have tropical climate with a

mean annual rainfall of 4687 mm (1996 − 2009, Department of Hydro-Met Services,

Bhutan). At elevations between 2000 and 3000 m, the annual precipitation is 3500 mm,

with summer temperatures reaching 29 °C and winter temperatures dropping to 3 °C in

December (Wangda et al., 2009).

Forests are most diverse in the mid-altitudes of the study area. Basal area

increases with altitude (Table 4.1). Soils are generally acidic, and the acidity increases

with altitude and decreases with depth. The lowest pH of 4.7 and the highest pH of 5.5

for the topsoil were found for CTBLF and TF, respectively. Conversely, the bulk density

(BD) decreases with altitude and increases with soil depth.

Table 4.1 Climate and forest characteristics along the altitudinal gradient grouped by

forest type

Forest

Type Altitude Species H´

BA m2

ha-1

MAT

(°C)

Tmax

(°C)

Tmin

(°C)

MA

RH %

RHmax

(%)

RH min

(%)

TF 317 - 900 33 2.96 17.8 22.9 28.9 12.5 73.0 92.2 55.7

STF 900 - 1870 54 3.47 30.1 15.4 21.3 7.0 85.3 96.5 69.7

WTBLT 1870 - 2450 42 3.21 41.1 13.4 20.6 3.9 84.9 97.8 64.7

CTBLF 2450 - 3000 47 3.14 46.3 10.9 17.0 -0.9 88.7 98.9 67.3

CTF 3000 - 3300 10 1.96 66.6 5.5 13.0 -5.8 85.2 99.2 56.5

(TF: Tropical forest, STF: Sub-tropical forest, WTBLF: Warm temperate broadleaf forest,

CTBLF: Cool temperate broadleaf forest and CTF: Cold temperate forest). H´: Shannon diversity

index, BA: basal area m2 ha-1, MAT: mean annual temperature, Tmax: maximum temperature, Tmin:

minimum, MARH (%): mean annual relative humidity, RHmax (%): mean annual maximum


Page | 79

relative humidity, RHmin (%): mean annual minimum relative humidity. MAT, Tmax and Tmin were

measured in 2012, whereas MARH, RHmax and RHmin (%) were recorded in 2011.

4.2.2 Plant sampling

Sampling was conducted at 150 m (biomass) and 75 m (soil) altitudinal

intervals. Altogether 20 vegetation inventory plots and 40 soil profiles were surveyed. To

determine the forest composition and basal area of the forest a 30 m × 30 m plot was

established to conduct the overstorey tree inventory. Within each inventory plot, the

understorey vegetation (diameter at 1.3 m height < 2.5 cm), overstorey litter (O_lt) and

branch deadwood (O_bdw) < 5 cm diameter were collected from three 1 m × 1 m plots.

Additionally, 144 trees (DBH > 10 cm) were harvested along the transect and a wood

core (O_w) and 5 leaf samples each from the sunlit and shaded canopy were collected.

Overstorey leaves (O_l) were pressed in a plant press and understorey biomass

was segregated into leaves (U_l), branches (U_b) and stem wood (U_w) and bagged. All

biomass samples were transferred to the laboratory and oven-dried at 60 °C. Woody

samples (branch, stem and dead wood) were sliced into thin pieces with a knife to make

them suitable for grinding. A Philips HL 1606/00 mixer-grinder was used to coarsely

grind each of the biomass samples. A representative aliquot was taken from each of the

coarse samples and finely ground using a Retsch MM400 Mixer Mill (RETSCH GmbH,

Haan, Germany) for isotope and elemental analyses.

4.2.3 Soil sampling

Soil samples were collected from soil profiles (0 – 100 cm) at altitudinal

intervals of 75 m. Soils sampled were composited into four depth categories i.e. 0 – 10

cm; 10 – 30 cm; 30 – 60 cm and 60 – 100 cm and oven dried at 40 °C for 48 h. Dried soil

samples were ground, passed through a 2 mm sieve to separate pebbles and stones and all

proportions weighed. Representative soil samples were ground using a Fritsch

Pulverisette 2 Mortar Grinder Mill (RETSCH GmbH, Haan, Germany), sieved with a 53

µm sieve and then used for isotope and elemental analyses.


Page | 80

4.2.4 Stomatal measurements and calculation of gw max

Leaf stomatal dimensions and densities were determined for sun and shade

leaves of 49 individuals of Persea sp. under a light microscope (Leica DM 2500M,

Germany) for the 3 mid-altitude forests (STF, WTBLF and CTBLF). Additionally, leaf

stomatal dimensions and densities were determined for sun-exposed leaves of 63

randomly chosen tree species from all forest types under the Jeol Neoscope table top

scanning electron microscope. The images were analysed using LAS V4.0 and image J

software (Schneider et al., 2012). Only the abaxial surface of the leaves had stomata and

was considered for image analysis. At least five measurements of guard cell length, pore

length, guard cell width and number of stomata at 400 × magnification were recorded.

Maximum stomatal conductance to water vapour (gw max) in moles per meter squared per

second (mol m-2 s-1) was estimated using the following equation (Franks et al., 2009;

Franks and Farquhar, 2001):

gwmax = d Damax

v�(l+π2 � amaxπ �

Equation 4.1

where d is the diffusivity of water in air (m2 s-1); D is the density of stomata (m-

2); amax is the mean maximum stomatal pore area (m-2); v is the molar volume of air (m3

mol-1); l is the depth of the stomatal pore (m) approximated as the W/2 for fully inflated

guard cells (Franks and Farquhar, 2007) and π is the mathematical constant (3.142).

4.2.5 Isotope and elemental analysis

δ13C, δ15N and C and N concentrations in the soil and plant material was

determined on a Thermo Delta V isotope ratio mass spectrometer coupled to ConfloIV

and FlashHT peripherals (Thermo Fisher, Bremen, Germany). Approximately 3 mg of

plant and 3 – 30 mg of soil material were weighed into tin cups, folded and combusted at

1000 °C. δ13C values are expressed in ‰ on the VPDB (Vee Pee Dee Belemnite) scale.

Analytical precision was better than 0.11‰ (δ13C), 0.12‰ (δ15N), 0.15% (C) and 0.18%

(N).


Page | 81

4.2.6 Carbon and nitrogen isotopic enrichment with depth

To gauge C and N isotopic enrichment with soil depth for each of the forest

types, δ13C and δ15N was regressed with their respective log transformed C and N

concentration (g kg-1). The slopes are an indication of the rate of change in 13C (Garten et

al., 2000) and 15N with organic matter decomposition.

4.2.7 Statistical analysis

IBM SPSS 21 was used to perform all statistical analyses, including one way

ANOVA to compare δ13C, C %, δ15N, N % and stomatal measurements in the sunlit and

shaded leaves within each forest type; δ13C and δ15N of overstorey leaves, litter, branch

deadwood and understorey leaves, stem wood and branch wood from different forest

types along the altitudinal gradient; differences between δ13C and δ15N in different soil

depths and forest types along the altitudinal gradient. A regression analysis was

performed to determine the relationship between δ13C in the soil and δ13C in the leaf litter

and deadwood along the altitudinal gradient. A Pearson product moment correlation

between the δ13C of soils and δ13C of leaf litter and dead wood was carried out to test the

dependency of these variables. Additionally, the slopes of the linear regression of C

isotope and logarithm of C concentration for each forest types were derived.

4.3 Results

4.3.1 Biomass carbon and nitrogen isotope trends with forest types

δ13C in all overstorey biomass components (branch deadwood, litter, overstorey

sunlit and shade leaves as well as tree wood) were significantly different among forest

types, whereas the δ13C values of understory biomass for the forest types were broadly

similar (Table 4.2). Only δ15N in the litter and overstorey tree leaves showed significant

variation among the forest types, whereas δ15N of branch deadwood and understorey

biomass were comparable (Table 4.2). Average δ13C of overstorey sunlit and shade leaves

were enriched compared to overstorey woody components (branch deadwood and trunk


Page | 82

wood) for all forest types (with the exception of branch deadwood in TF) and enrichment

increased with altitude of the forest (Fig. 4.1). Additionally, overstorey tree wood was the

most enriched biomass component for all forest types (Fig. 4.1). In contrast, litter δ13C

had a more variable response, with the lowest (TF and STF) and highest altitude (CTF)

forests being more depleted than the overstorey.

Table 4.2 Carbon and nitrogen isotopes (Mean ± S.D) in different biomass categories for

different forest zones. O_bdw = overstorey tree branch deadwood, O_lt = overstorey litter

O_l = average of overstorey sun and shade tree leaves, O_w = overstorey wood, U_b =

understorey branch, U_l = understorey leaf, U_w = understorey wood. TF: Tropical

forest, STF: Sub-tropical forest, WTBLF: Warm temperate broadleaf forest, CTBLF:

Cool temperate broadleaf forest and CTF: Cold temperate forest. Forest Zone Overstorey biomass Understorey biomass

δ 13C O_bdw δ 13C O_lt δ 13C O_l δ 13C O_w δ 13C U_b δ 13C U_l δ 13C U_w TF -29.1ab ± 1.1 -29.7ab ± 0.8 -28.7a ± 1.8 -27.92a ± 1.0 -31.0a ± 1.6 -31.3a ± 1.7 -29.9a ± 1.2 STF -29.5a ± 1.1 -30.2a ± 0.5 -29.8b ± 1.5 -27.83a ± 1.7 -31.5a ± 1.2 -31.7a ± 1.2 -31.2b ± 0.5 WTBLF -29.3ab ± 1.0 -29.7ab ± 0.4 -29.8b ± 1.4 -28.05a ± 1.5 -30.3a ± 0.8 -30.8a ± 1.2 -30.4ab ± 1.1 CTBLF -28.3ab ± 1.0 -29.1b ± 0.6 -29.4b ± 1.9 -27.80a ± 1.7 -30.8a ± 1.4 -31.0a ± 0.9 -30.6ab ± 0.9 CTF -27.4 b ± 0.7 -28.6b ± 0.5 -28.3a ± 1.5 -25.57b ± 1.2 -30.2a ± 1.7 -30.0a ± 0.4 -30.1ab ± 1.7 δ 15N O_bdw δ 15N O_lt δ 15N O_l δ 15N U_b δ 15N U_l δ 15N U_w TF -2.5a ± 0.6 -1.9a ± 0.7 -0.6a ± 1.3 - -2.4a ± 0.7 -1.1a ± 0.8 -2.7a ± 0.9 STF -2.6a ± 1.3 -2.1b ± 1.6 -1.8bd ± 1.1 - -3.2a ± 1.4 -1.7a ± 1.7 -3.4ab ± 1.5 WTBLF -2.7a ± 0.9 -2.1ab ± 0.7 -1.3b± 1.2 - -2.7a ± 0.7 -1.9a ± 0.9 -2.6a ± 0.4 CTBLF -3.5a ± 0.9 -3.4b ± 1.5 -2.7c ± 1.7 - -4.8b ± 0.9 -3.9b ± 0.7 -4.6b ± 1.6 CTF -3.5a ± 1.0 -2.9ab ± 0.8 -1.9d ± 2.0 - -3.2ab± 0.5 -1.4a ± 0.5 -3.5ab ± 0.7

a Means (column) followed by the same letters are not significantly different at P = 0.05.


Page | 83

Fig. 4.1 Differences of δ13C in overstorey leaves (average of sun and shade) to other

biomass components (O_w: overstorey wood, O_bdw: overstorey branch deadwood and

O_lt: overstorey litter; expressed as δ13C leaf – biomass) in each forest type (TF: Tropical

forest, STF: Sub-tropical forest, WTBLF: Warm temperate broadleaf forest, CTBLF:

Cool temperate broadleaf forest and CTF: Cold temperate forest).

4.3.2 Biomass carbon and nitrogen isotope trends with altitude

δ13C of the overstorey biomass components showed a curvilinear relationship with

altitude. δ13C of litter (Fig. 4.2a), branch dead wood and tree wood (Fig. 4.2b) decreased

with increasing altitude up to ~1,500 m a.s.l.. δ13C of canopy sun and shade leaves (Fig.

4.2c) decreased as altitude increased to ~1800 m a.s.l. and thereafter increased with

altitude up to 3300 m a.s.l.. In contrast, there was no observable trend in δ13C in

understorey biomass (leaves, branch, and wood) with altitude (Fig. 4.2d−f).


Page | 84

Fig. 4.2 Altitudinal trend of δ13C in overstorey (a) sun (black symbols and solid line) and

shade (grey symbols and dotted line) leaves, b) litter, c) branch deadwood (black symbols

and solid line) and tree trunk wood (grey symbols and dotted line), and understorey

biomass d) leaves, e) branch and f) stem wood.

The δ15N value of understorey vegetation and all overstorey biomass

components (litter, branch deadwood and sun and shade leaves) decreased significantly

with increasing altitude. δ15N values of overstorey sun and shade leaves and litter

decreased by 0.7 ‰ (Fig. 4.3a&b) and of branch deadwood by 0.5‰ (Fig. 4.3c) for every


Page | 85

1000 m increase in altitude. There was no significant difference in δ 15N between sun and

shade canopy leaves along the altitudinal gradient.

The δ15N values of understorey leaves decreased by 0.8‰ (Fig. 4.3d), and of

branch wood by 0.7‰ (Fig. 4.3e) and stem wood by 0.5‰ (Fig. 4.3f), for every 1000 m

increase in altitude.

Fig. 4.3 Altitudinal trend of δ15N values in overstorey (a) sun (black symbols and solid

line) and shade (grey symbols and dotted line) leaves, b) litter, c) branch deadwood and


Page | 86

understorey (d) leaves, e) branch and f) stem wood biomass components. Solid lines are

drawn when relationships are significant at P = 0.05

4.3.3 Tree stomatal density and conductance with altitude and forest types

In order to evaluate the effect of stomatal characteristics on δ13C, we calculated

maximal stomatal conductance from stomatal dimensions and density for the tree genus

Persea, which grew at STF, WTBLF and CTBLF. There was no significant difference in

gwmax between the three forest types (or altitudes), although the average length of the

guard cell pores was larger, (t 47 = −2.18, P = 0.034) in shaded leaves than in sunlit

leaves. The density of stomata (t 47 = 2.8, P = 0.007) however, was greater in the sunlit

leaves than in shade leaves (Supplement 4.1).

We then estimated gwmax for all tree species. There was no difference in pore

length, stomatal density and gwmax (Supplement 4.2) among all forest types along the

altitudinal gradient. Although statistically non-significant, gwmax increased with altitude

and levelled off at about 1800 m a.s.l and then decreased thereon with increasing altitude

(Supplement 4.3), opposite to the δ13C trend. This resulted in a linear (albeit

insignificant) relationship between the two parameters with a shallow negative slope (Fig.

4.4).


Page | 87

Fig. 4.4 Relationship of maximum potential stomatal conductance with δ13C in sunlit

overstorey leaves.

4.3.4 Soil carbon and nitrogen isotope trends with forest type, altitude and soil depth.

Average soil δ13C was significantly different among forest types along the

altitudinal gradient (F4, 162 = 17.52, P < 0.001). Soil δ13C decreased at all soil depths as

altitude increased up to 1800 m a.s.l.. At greater elevations, soil δ13C increased with

altitude to a maximum of 3300 m a.s.l.. Rates of change in δ13C were: 1.7‰ km-1 for 0 –

10 cm 1.5‰ km-1 for the 10 – 30 cm, 1.4‰ km-1 for 30 – 60 cm, and 1.8‰ km-1 for the

60 – 100 cm soil depth (Fig. 4.5a). This trend is similar to changes in δ13C of litter,

branch deadwood and overstorey tree leaves and wood with altitude. Additionally, soil

δ13C increased with depth (0 – 100 cm) for all forest types (F3, 162 = 15.29, P = 0.001, Fig.

4.5a). Differences between surface layer (0 – 10 cm) and the deepest layers (60 – 100 cm)

were greatest in the TF (2.11‰) and least in the CTF (0.93‰).


Page | 88

Fig. 4.5 δ13C ± S.D (a) and δ 15N ± S.D (b) at different soil depths for forest types in the

study area. Bars with same letters indicate that there is a non- significant difference

across forest types for the same category of soil depth (P = 0.05). TF = Tropical forest,

STF = Sub-tropical forest, WTBLF = Warm temperate broadleaf forest, CTBLF = Cool

temperate broadleaf forest, CTF = Cold temperate forest.

Soil δ15N did not vary significantly along the altitudinal gradient (F4, 162 = 1.317,

P = 0.266, Fig. 4.5b). Due to biomass δ15N values decreasing with altitude, differences

between plant δ15N values and soil δ15N values (Δδ15Nplant-soil, (Amundson et al., 2003)

increased with altitude (Supplement 4.4). Similar to δ13C in the soil, there was a

significant increase in δ15N with soil depth from surface to 100 cm (F3, 162 = 22.185, P <

0.001, Fig. 4.5b) for all forest types. In contrast to δ13C, differences between δ15N values

for surface (0 − 10 cm) and deepest soil layers (60 − 100 cm) were greatest at highest

altitudes (CTF; 3.13‰).

4.3.5 C:N ratio in the soil along the altitudinal gradient

Soil C:N ratio varied between forest types (F4, 162 = 56.75, P < 0.001) as well as

between soil depths (F3,162 = 2.82, P = 0.041). Altitude accounted for 50 – 60% of

increases in C:N ratio at different soil depths (Fig. 4.6). There was no change in the C:N

ratio with soil depth for lower altitude forests (TF and STF), but C:N increased between

these two sites. C:N ratios increased (although statistically insignificant) with depth for


Page | 89

mid and high-elevation forests (WTBLF, CTBLF and CTF). Soil C:N ratios for surface

horizons (0 − 10 cm) ranged between 9 and 15 as compared to 8 − 17.5 for the 60 – 100

cm soil depth. The average C:N ratio in all soils at all depths, increased by ~ 2.2 for every

1000 m increase in altitude.

Fig. 4.6 Soil carbon to nitrogen ratio at different depths and under different forest types

along the altitudinal gradient. TF = Tropical forest, STF = Sub-tropical forest, WTBLF =

Warm temperate broadleaf forest, CTBLF = Cool temperate broadleaf forest, CTF = Cold

temperate forest. Temperatures are mean annual temperatures for the forest types.

4.3.6 Relationship of carbon and nitrogen isotopes to total C and N concentrations in

soil

Log C concentration and soil δ13C values were negatively correlated for all

forest types, except for CTF at the highest altitude and the slope of the regression

decreased with increasing altitude of the forest (Fig. 4.7a−d). For CTF at the highest

altitude, δ13C values in the soil remained constant irrespective of the change in log C

concentrations in the soil (Fig. 4.7e).


Page | 90

Similar to C, log N concentrations (g kg-1) and δ15N in the soil were negatively

correlated for all forest types along the altitudinal gradient. In contrast to C however, the

slope of the regression between log N concentrations (g kg-1) and δ15N values increased

with increasing altitude of the forest ( Fig. 4.7f−i).

Fig. 4.7 Relationship between δ13C and log transformed C concentration (g kg-1, a − e),

and δ15N and log transformed N concentration (g kg-1,f − j) in soils with depth (0 − 100

cm). TF = Tropical forest, STF = Sub-tropical forest, WTBLF = Warm temperate

broadleaf forest, CTBLF = Cool temperate broadleaf forest, CTF = Cold temperate forest.


Page | 91

4.3.7 Correlations between soil and biomass C and N isotopes and soil properties

Soil C and N isotope signatures were mostly significantly related to overstorey

components (Table 4.3). Soil δ13C values were negatively correlated with soil C and N

concentrations and CEC. Altitude and temperature were significantly correlated with C

and N isotopes in different biomass components, soil C and N concentrations, soil

properties like CEC, C:N ratio and BD, but not soil C and N isotopes. Soil C content for

the entire soil profile was not correlated with clay content; however, when C

concentration and clay content were considered separately for each of the soil depth

categories, correlation coefficients increased from shallow to deeper depths (statistically

significant, Supplement 4.5).


Page | 92

Table 4.3 Pearson correlation coefficients between carbon and nitrogen isotopes and concentrations in soil (averages of 0 – 100 cm)

and C and N isotopes in biomass. Significant correlations are marked with * (<0.05) and ** (<0.01). Alt Temp Soil

δ13 C Soil δ 15N

δ13 C Wood

δ13 C leaf C

δ 15N leaf C

δ13 C Br

δ 15N Br

δ13 C Dw

δ 15N Dw

δ13 C Fo

δ 15N Fo

δ13 C Lt

δ 15N Lt

C (g kg_1)

N (g kg_1)

CEC CN ratio

BD EC_ Sand (%)

Silt (%)

Clay (%)

Alt 1

Temp -.93** 1

Soil δ13 C .12 -.10 1

Soil_ δ 15N .05 -.00 .61** 1

δ13 C wood

.54** -.51**

.47** .22* 1

δ13 C leaf C

-.10 .20 .31** .10 .50** 1

δ15N leaf C

-.55** .54** .00 .10 .03 .48** 1

δ13 C Br .29* -.23* .10 .10 .17 -.24* -.37**

1

δ 15N Br -.49** .55** .00 .20 -.22 .16 .53** .29** 1

δ13 C Dw .52** -.41**

.45** .27* .58** .27* -.32**

.20 -.20 1

δ 15N Dw -.46** .50** .00 .26* -.11 .13 .56** .10 .77** -.22 1

δ13 C Fo .25* -0.2 .10 .10 .07 -.26* -.28* .88** .35** .28* .10 1

δ 15N Fo -.52** .57** .20 .20 .01 .23* .55** .10 .90** -.06 .82** .15 1

δ13 C Lt .61** -.38**

.33** .20 .49** .28* -.38**

.20 -0.2 .78** -.23* .24* -.10 1

δ 15N Lt -.50** .54** -.00 .22* -.10 .13 .51** .20 .79** -.21 .93** .14 .87** -.20 1

C (g kg_1) .62** -.57**

-.39**

-.38**

.31** -.09 -.33**

.23* -.33**

.24* -.29**

.15 -.38**

.30** -.27* 1

N (g kg_1) .45** -.41**

-.46**

-.43**

.16 -.12 -.22* .20 -.20 .11 -0.2 .13 -.28* .18 -.20 .94** 1

CEC .63** -.57**

-.28* -.30**

.32** -.07 -.40**

.10 -.53**

.40** -.42**

.09 -.53**

.36** -.45**

.89** .80** 1

CN ratio .74** -.69**

.10 .20 .46** -.01 -.37**

.10 -.46**

.38** -.44**

.14 -.45**

.41** -.46**

.32** .02 .41** 1

BD -.75** .72** .26* .20 -.30**

.19 .41** .00 .50** -.34**

.38** .03 .48** -.41**

.43** -.81** -.72** -.81**

-.50** 1

EC_ .20 -.10 -.34**

-.52**

.09 .01 -.25* .00 -0.2 .18 -0.2 -.10 -0.1 .22* -0.2 .53** .53** .57** -.00 -.47** 1

Sand (%) -.00 .00 .10 -.00 .21 .27* .12 .23* .31** -.06 .10 .16 .32** .11 .26* -.01 -.05 -.31**

-.00 .27* -.20 1

Silt (%) .03 -.00 -.24* -.00 -.29**

-.24* -.10 -.39**

-.29* -.07 -.10 -.29**

-.30**

-.11 -.25* .10 .06 .24* .05 -.28* .21 -.85**

1

Clay (%) .00 -.00 0.1 .29** .02 -.17 -.10 .10 -.20 .21 -.00 .12 -.20 -.04 -.10 -.02 .01 .23* -.00 -.10 .05 -.64**

.14 1

leaf C = Sun lit leaves, Br = understorey vegetation branch wood, Dw = Deadwood, Fo = understorey vegetation leaves, Lt = litter on the forest floor, CEC =

cation exchange capacity, C:N ratio = soil C:N ratio, BD = soil bulk density, EC = soil electric conductivity.


Page | 93

4.4 Discussion

4.4.1 Biomass and soil carbon isotope trends with forest type and altitude

δ13C in biomass along the altitudinal gradient was clearly related to light. For

instance, δ13C for understorey components (all forest types and altitudes), were

consistently more depleted than overstorey leaves, branch deadwood and litter (Table

4.2), arguably as a result of light limitation (Gessler et al., 2004). In contrast, δ13C of

overstorey biomass components varied with altitude and were greatest in mid-altitude

forests, similar to studies by Luo et al. (2006) and Zhao et al. (2008) in China. Some of

these trends are explained by greater water availability (see also Wang et al. (2010))

and/or higher stomatal conductance at mid-altitudes. In our study, however, water

limitation seems an unlikely cause of variation in δ13C of overstorey trees, since rainfall

is abundant (rainfall range = 3500 to 4700 mm) and greatest at low altitudes where δ13C

was less negative. Instead, we propose that fog, which is common in Bhutan (Gratzer et

al., 1999), and especially persistent between lower and mid-altitudes, provides a

significant light limitation. These fogs gradually disappear as elevation increases, and

are rare at the highest altitudes. Fog is correlated with relative humidity (Syed et al.,

2012) and mean annual relative humidity for study sites (Table 4.1) corresponded with

observed patterns of fog. Reduced δ13C can be due to reduced average maximum rates of

net photosynthesis (Amax) at constant stomatal conductance (g), increased g at constant

Amax, or a combination of these two processes (Farquhar et al., 1982). While gw max in

this study showed a curvilinear relationship with altitude (some of which may have been

influenced by higher plant diversity at mid-altitudes which increases water use and

stomatal conductance (Caldeira et al., 2001), the flat linear relationship between δ13C

and gw max suggests that most of the variation in δ13C of the overstorey components is

due to variation in Amax, driven by light. From 1800 m a.s.l. δ13C of overstorey

components increased at a rate of c. 1‰ per km comparative to other studies, e.g. 1.4 –

1.7‰ per km (Vitousek et al., 1990), 1.33‰ per km (Marshall and Zhang, 1994), 0.9‰


Page | 94

per km (Wang et al., 2010) or 1.9 – 2.5‰ for conifers (Warren et al. (2001). These

trends are attributed to decreased C isotope discrimination during photosynthesis

associated with greater carboxylation efficiency with increasing altitude (Körner et al.,

1988; Körner et al., 1991). Moreover, cooler temperatures can influence δ13C via the

slowing of sap flow, thereby increasing gradients in water potential between soil and

leaf causing partial stomatal closure (corroborated by the decrease in gw max for higher

altitudes here) and increasing δ13C (e.g. Meinzer et al. (1992).

The δ13C values of heterotrophic tissues (overstorey wood and branch

deadwood) were more enriched than that of overstorey leaves (see Fig. 4.1) which is a

phenomenon widely found in C3 plants and was reported as early as 1982 by Leavitt and

Long (1982). This enrichment can be influenced by loading of photosynthates into the

phloem, transient starch accumulation and hydrolysis at night, as well as during

basipetal transport of carbohydrates (Cernusak et al., 2009; Gessler et al., 2009; Gessler

et al., 2014). Since the offset was smaller for branch deadwood compared to tree wood

(Fig. 4.1), we argue that this is evidence of fractionation during basipetal transport.

Additionally, the magnitude of the difference increased with altitude at a rate of about

−0.11‰, which is less than half the “temperature coefficient” of −0.27‰ °C-1 found by

(Leavitt and Long, 1982).

The δ13C values of litter were more depleted than overstorey leaves for the TF,

STF and CTF. For the lower and warmer forests, litter decomposes faster (Lloyd and

Taylor, 1994) accompanied by the loss of 13C (Fernandez et al., 2003). Although branch

deadwood and litter on the forest floor derives mostly from overstorey trees (Medina and

Minchin, 1980; Prescott, 2002), 13C depleted understorey vegetation made up most of

the litter and potentially some of the branch dead wood at the highest elevation

(broadleaf species rather than overstorey conifers), leading to a smaller (dead wood) or

negative (litter) offset to overstorey leaf δ13C. Despite relatively more contribution of the

understorey to litter at the highest site, the important contribution of the canopy to soil C

along the whole transect (see Peri et al., 2012) was demonstrated by the curvilinear


Page | 95

pattern of biomass δ13C with altitude which translated to a similar δ13C pattern in the soil

at all depths. Furthermore, higher temperature and greater microbial diversity at lower

altitudes (Biasi et al., 2005) favoured by more neutral soils (Fierer and Jackson, 2006),

facilitates the breakdown of C fractions that are depleted in 13C, while fractionation

during decomposition increases with decreasing temperatures (Schaub and Alewell,

2009), both processes leaving the soil enriched in 13C. Soil fungi discriminate against the

heavier stable isotopes during degradation of C compounds (Henn and Chapela, 2000)

and soil organic matter becomes relatively enriched in 13C (Nadelhoffer and Fry, 1988)

as decomposition progresses (e.g. with soil depth). C concentration, CEC and C:N ratio

increased with altitude and with depth at higher altitudes, indicating decomposition was

likely faster at lower altitudes and at shallower depths. When regressing soil δ13C to

logarithm of C concentration, which is indicative of abundance of 13C related to the

degree of organic matter decomposition (Garten, 2006), highest altitude forests (CTF

and CTBLF) had the shallowest slopes, confirming that soil C turnover is slowest for the

highest altitude forest. Additionally, the smaller difference between surface soil (0 – 10

cm) and deeper soil (60 – 100 cm) at the highest altitude suggests that there is little C

lost via leaching in deeper layers. Increasing correlation of C and clay content with

altitudes at depths of 30 – 100 cm may be an indication of organo-mineral stabilisation

(Jones and Singh, 2014; Torn et al., 1997) of C in deep soil layers at higher altitudes.

4.4.2 Biomass and soil nitrogen isotopes trends along the altitudinal gradient.

The δ15N values of overstorey leaves, litter, branch deadwood as well as

understorey biomass decreased linearly with increasing altitude due to decreasing MAT

(Amundson et al., 2003). Lower biomass δ15N with increasing altitude may be due to a

‘less open’ N cycle (i.e. little N is lost) at higher altitude forest sites (Karolien et al.,

2013). This decreasing trend in δ15N values was absent in the soil (average of all depths)

and suggests that total soil N was not representative of the N taken up by plants. At

lower altitudes, where litter decompostion is faster and hence 14N can be preferentially

lost, the remaining N taken up by plants is higher in 15N (Amundson et al., 2003;


Page | 96

Delwiche et al., 1979) or may originate from deeper soils with higher δ15N values. We

observed an increase in the difference between plant and soil δ15N with altitude (Δδ

15Nplant-soil) which may have been associated with a shift of N-source from NO3− in

tropical forests to organic/NH4+ N in the cooler high altitude forests (Amundson et al.,

2003). Additionally, increasing altitude and decreasing temperature can cause a shift in

the associated mycorrhiza from arbuscular to ecto or ericoid mycorrhizal species and

associated decreases in δ15N (Craine et al., 2009).

In soils, δ15N increased with altitude only in the deeper layers, resulting in the

largest difference in δ15N between the top 10 cm and the deepest soil at the highest

altitude forest (CTF). More progressed mineralisation as well as decreased biological

activity at higher altitudes due to lower temperatures increases δ15N (Dijkstra et al.,

2006; Lloyd and Taylor, 1994; Mariotti et al., 1980) and may have contributed to higher

δ15N values in deeper soils. In contrast to C, 14N may have been lost due to leaching

and/or N having a longer residence time in deeper high altitude forest soils (Martinelli et

al., 1999), since the largest negative slope of the regression of soil δ15N to logarithm of

N concentration was found at the highest altitude. The contrasting trends of stable

isotopes in relation to concentration for carbon and nitrogen suggests that N dynamics

are decoupled from the C dynamics in the soil, i.e. there is potential stabilisation of C in

deeper layers, while N is lost to leaching.

Stable carbon isotopes verified that overstorey biomass contributed the majority

of C to soils along this altitudinal gradient. A curvilinear trend of overstorey δ13C, likely

a result of varying light limitations, translated directly into soil δ13C at all depths. Plant

physiological functions, such as photosynthesis and stomatal conductance, therefore

appear to exert a strong influence in determining soil δ13C patterns along altitudinal

gradients compared to decomposition processes. With depth, isotopic gradients of δ13C

and δ15N (supported by measures of e.g. C:N ratio) indicated possible slowing of

decomposition and loss of N in deeper soils at higher altitudes. Smaller offsets in δ13C

between shallow and deep soils at high altitude point to potential stabilisation of C in


Page | 97

deep soils by physical or chemical processes which are an area of further investigation.

However, with the progression of global warming, C and N stocks in shallow soils,

especially in high altitude forests, will be highly vulnerable to losses due to increased

decomposition.


Page | 98

References

Amundson, R., Austin, A.T., Schuur, E.A.G., Yoo, K., Matzek, V., Kendall, C.,

Uebersax, A., Brenner, D., Baisden, W.T., 2003. Global patterns of the isotopic

composition of soil and plant nitrogen. Global Biogeochemical Cycles 17(1),

1031.

Austin, A.T., Sala, O.E., 1999. Foliar δ15N is negatively correlated with rainfall along

the IGBP transect in Australia. Functional Plant Biology 26(3), 293-298.

Bai, E., Boutton, T.W., Liu, F., Wu, X.B., Hallmark, C.T., Archer, S.R., 2012. Spatial

variation of soil 13C and its relation to carbon input and soil texture in a

subtropical lowland woodland. Soil Biology and Biochemistry 44(1), 102-112.

Beerling, D.J., Heath, J., Woodward, F.I., Mansfield, T.A., 1996. Drought-CO2

interactions in trees: Observations and mechanisms. New Phytologist 134(2),

235-242.

Biasi, C., Rusalimova, O., Meyer, H., Kaiser, C., Wanek, W., Barsukov, P., Junger, H.,

Richter, A., 2005. Temperature-dependent shift from labile to recalcitrant carbon

sources of arctic heterotrophs. Rapid Communications in Mass Spectrometry

19(11), 1401-1408.

Bird, M.I., Chivas, A.R., Head, J., 1996. A latitudinal gradient in carbon turnover times

in forest soils. Nature 381(6578), 143-146.

Bird, M.I., Haberle, S.G., Chivas, A.R., 1994. Effect of altitude on the carbon-isotope

composition of forest and grassland soils from Papua New Guinea. Global

Biogeochemical Cycles 8(1), 13-22.

Caldeira, M.C., Ryel, R.J., Lawton, J.H., Pereira, J.S., 2001. Mechanisms of positive

biodiversity–production relationships: Insights provided by δ13C analysis in

experimental Mediterranean grassland plots. Ecology Letters 4(5), 439-443.


Page | 99

Cernusak, L.A., Tcherkez, G., Keitel, C., Cornwell, W.K., Santiago, L.S., Knohl, A.,

Barbour, M.M., Williams, D.G., Reich, P.B., Ellsworth, D.S., Dawson, T.E.,

Griffiths, H.G., Farquhar, G.D., Wright, I.J., 2009. Why are non-photosynthetic

tissues generally 13C enriched compared with leaves in C3 plants? Review and

synthesis of current hypotheses. Functional Plant Biology 36(3), 199-213.

Craine, J.M., Elmore, A.J., Aidar, M.P., Bustamante, M., Dawson, T.E., Hobbie, E.A.,

Kahmen, A., Mack, M.C., McLauchlan, K.K., Michelsen, A., 2009. Global

patterns of foliar nitrogen isotopes and their relationships with climate,

mycorrhizal fungi, foliar nutrient concentrations, and nitrogen availability. New

Phytologist 183(4), 980-992.

Curtis, P.S., Hanson, P.J., Bolstad, P., Barford, C., Randolph, J.C., Schmid, H.P.,

Wilson, K.B., 2002. Biometric and eddy-covariance based estimates of annual

carbon storage in five eastern North American deciduous forests. Agricultural

and Forest Meteorology 113(1–4), 3-19.

Delwiche, C.C., Zinke, P.J., Johnson, C.M., Virginia, R.A., 1979. Nitrogen isotope

distribution as a presumptive indicator of nitrogen fixation. Botanical Gazette

140. Supplement: Symbiotic Nitrogen Fixation in Actinomycete-Nodulated

Plants.March 1979. The University of Chicago Press, S65-S69.

Dieleman, W.I.J., Venter, M., Ramachandra, A., Krockenberger, A.K., Bird, M.I., 2013.

Soil carbon stocks vary predictably with altitude in tropical forests: Implications

for soil carbon storage. Geoderma 204–205, 59-67.

Dijkstra, P., Ishizu, A., Doucett, R., Hart, S.C., Schwartz, E., Menyailo, O.V., Hungate,

B.A., 2006. 13C and 15N natural abundance of the soil microbial biomass. Soil

Biology and Biochemistry 38(11), 3257-3266.

Du, B., Kang, H., Pumpanen, J., Zhu, P., Yin, S., Zou, Q., Wang, Z., Kong, F., Liu, C.,

2014. Soil organic carbon stock and chemical composition along an altitude


Page | 100

gradient in the Lushan Mountain, subtropical China. Ecological Research 29(3),

433-439.

Farquhar, G., O'Leary, M., Berry, J., 1982. On the relationship between carbon isotope

discrimination and the intercellular carbon dioxide concentration in leaves.

Functional Plant Biology 9(2), 121-137.

Farquhar, G., Richards, R., 1984. Isotopic composition of plant carbon correlates with

water-use efficiency of wheat genotypes. Functional Plant Biology 11(6), 539-

552.

Fernandez, I., Mahieu, N., Cadisch, G., 2003. Carbon isotopic fractionation during

decomposition of plant materials of different quality. Global Biogeochemical

Cycles 17(3), 1-11.

Fierer, N., Jackson, R.B., 2006. The diversity and biogeography of soil bacterial

communities. Proceedings of the National Academy of Sciences of the United

States of America 103(3), 626-631.

Fisichelli, N.A., Stefanski, A., Frelich, L.E., Reich, P.B., 2015. Temperature and leaf

nitrogen affect performance of plant species at range overlap. Ecosphere 6(10),

1-4.

France-Lanord, C., Derry, L.A., 1994. δ13C of organic carbon in the Bengal Fan: Source

evolution and transport of C3 and C4 plant carbon to marine sediments.

Geochimica et Cosmochimica Acta 58(21), 4809-4814.

France-Lanord, C., Derry, L.A., 1997. Organic carbon burial forcing of the carbon cycle

from Himalayan erosion. Nature 390(6655), 65-67.

Franks, P.J., Drake, P.L., Beerling, D.J., 2009. Plasticity in maximum stomatal

conductance constrained by negative correlation between stomatal size and


Page | 101

density: an analysis using Eucalyptus globulus. Plant, Cell and Environment

32(12), 1737-1748.

Franks, P.J., Farquhar, G.D., 2001. The effect of exogenous abscisic acid on stomatal

development, stomatal mechanics, and leaf gas exchange in Tradescantia

virginiana. Plant Physiology 125(2), 935-942.

Franks, P.J., Farquhar, G.D., 2007. The mechanical diversity of stomata and its

significance in gas-exchange control. Plant Physiology 143(1), 78-87.

Garten, C.T., Jr., 2006. Relationships among forest soil C isotopic composition,

partitioning, and turnover times. Canadian Journal of Forest Research 36(9),

2157-2167.

Garten, C.T., Jr., Cooper, L.W., Iii, W.M.P., Hanson, P.J., 2000. Climate controls on

forest soil C isotope ratios in the Southern Appalachian Mountains. Ecology

81(4), 1108-1119.

Garten Jr, C.T., 2011. Comparison of forest soil carbon dynamics at five sites along a

latitudinal gradient. Geoderma 167–168, 30-40.

Gessler, A., Brandes, E., Buchmann, N., Helle, G., Rennenberg, H., Barnard, R.L., 2009.

Tracing carbon and oxygen isotope signals from newly assimilated sugars in the

leaves to the tree-ring archive. Plant, Cell and Environment 32(7), 780-795.

Gessler, A., Ferrio, J.P., Hommel, R., Treydte, K., Werner, R.A., Monson, R.K., 2014.

Stable isotopes in tree rings: towards a mechanistic understanding of isotope

fractionation and mixing processes from the leaves to the wood. Tree physiology.

doi: 10.1093/treephys/tpu040

Gessler, A., Rennenberg, H., Keitel, C., 2004. Stable isotope composition of

organiccompounds transported in the phloem of European Beech - Evaluation of


Page | 102

different methods of phloem sap collection and assessment of gradients in carbon

isotope composition during leaf-to-stem transport. Plant Biology 6(6), 721-729.

Gratzer, G., Rai, P.B., Glatzel, G., 1999. The influence of the bamboo Yushania

microphylla on regeneration of Abies densa in central Bhutan. Canadian Journal

of Forest Research 29(10), 1518-1527.

Handley, L.L., Austin, A.T., Stewart, G.R., Robinson, D., Scrimgeour, C.M., Raven,

J.A., Heaton, T.H.E., Schmidt, S., 1999. The 15N natural abundance (δ15N) of

ecosystem samples reflects measures of water availability. Functional Plant

Biology 26(2), 185-199.

Henn, M.R., Chapela, I.H., 2000. Differential C isotope discrimination by fungi during

decomposition of C3- and C4- derived sucrose. Applied and Environmental

Microbiology 66(10), 4180-4186.

Jones, E., Singh, B., 2014. Organo-mineral interactions in contrasting soils under natural

vegetation. Frontiers in Environmental Science 2, 1-15.

Karolien, D., Ilaria Del, G., Andrea, V., M Francesca, C., 2013. Assessment of soil C

and N stocks and fractions across 11 European soils under varying land uses.

Open Journal of Soil Science 2013.

Körner, C., Farquhar, G.D., Roksandic, Z., 1988. A Global survey of carbon isotope

discrimination in plants from high altitude. Oecologia 74(4), 623-632.

Körner, C., Farquhar, G.D., Wong, S.C., 1991. Carbon isotope discrimination by plants

follows latitudinal and altitudinal trends. Oecologia 88(1), 30-40.

Leavitt, S.W., Long, A., 1982. Evidence for 13C/12C fractionation between tree leaves

and wood. Nature 298(5876), 742-744.


Page | 103

Liu, X.Z., Wang, G., 2010. Measurements of nitrogen isotope composition of plants and

surface soils along the altitudinal transect of the eastern slope of Mount Gongga

in southwest China. Rapid Communications in Mass Spectrometry 24(20), 3063-

3071.

Lloyd, J., Taylor, J.A., 1994. On the temperature dependence of soil respiration.

Functional Ecology 8(3), 315-323.

Luo, J., Zang, R., Li, C., 2006. Physiological and morphological variations of Picea

asperata populations originating from different altitudes in the mountains of

southwestern China. Forest Ecology and Management 221(1–3), 285-290.

Mariotti, A., Pierre, D., Vedy, J.C., Bruckert, S., Guillemot, J., 1980. The abundance of

natural nitrogen 15 in the organic matter of soils along an altitudinal gradient

(Chablais, Haute Savoie, France). Catena 7(4), 293-300.

Marshall, J.D., Zhang, J., 1994. Carbon isotope discrimination and water-use efficiency

in native plants of the North-Central Rockies. Ecology 75(7), 1887-1895.

Martinelli, L.A., Piccolo, M.C., Townsend, A.R., Vitousek, P.M., Cuevas, E.,

McDowell, W., Robertson, G.P., Santos, O.C., Treseder, K., 1999. Nitrogen

stable isotopic composition of leaves and soil: Tropical versus temperate forests.


Medina, E., Minchin, P., 1980. Stratification of δ13C values of leaves in Amazonian rain

forests. Oecologia 45(3), 377-378.

Meinzer, F.C., Rundel, P.W., Goldstein, G., Sharifi, M.R., 1992. Carbon isotope

composition in relation to leaf gas exchange and environmental conditions in

Hawaiian Metrosideros polymorpha populations. Oecologia 91(3), 305-311.

Moles, A.T., Perkins, S.E., Laffan, S.W., Flores-Moreno, H., Awasthy, M., Tindall,

M.L., Sack, L., Pitman, A., Kattge, J., Aarssen, L.W., Anand, M., Bahn, M.,


Page | 104

Blonder, B., Cavender-Bares, J., Cornelissen, J.H.C., Cornwell, W.K., Díaz, S.,

Dickie, J.B., Freschet, G.T., Griffiths, J.G., Gutierrez, A.G., Hemmings, F.A.,

Hickler, T., Hitchcock, T.D., Keighery, M., Kleyer, M., Kurokawa, H.,

Leishman, M.R., Liu, K., Niinemets, Ü., Onipchenko, V., Onoda, Y., Penuelas,

J., Pillar, V.D., Reich, P.B., Shiodera, S., Siefert, A., Sosinski, E.E.,

Soudzilovskaia, N.A., Swaine, E.K., Swenson, N.G., van Bodegom, P.M.,

Warman, L., Weiher, E., Wright, I.J., Zhang, H., Zobel, M., Bonser, S.P., 2014.

Which is a better predictor of plant traits: temperature or precipitation? Journal

of Vegetation Science 25(5), 1167-1180.

Mueller, K.E., Hobbie, S.E., Chorover, J., Reich, P.B., Eisenhauer, N., Castellano, M.J.,

Chadwick, O.A., Dobies, T., Hale, C.M., Jagodziński, A.M., Kałucka, I.,

Kieliszewska-Rokicka, B., Modrzyński, J., Rożen, A., Skorupski, M., Sobczyk,

Ł., Stasińska, M., Trocha, L.K., Weiner, J., Wierzbicka, A., Oleksyn, J., 2015.

Effects of litter traits, soil biota, and soil chemistry on soil carbon stocks at a

common garden with 14 tree species. Biogeochemistry 123(3), 313-327.

Nadelhoffer, K.J., Fry, B., 1988. Controls on natural nitrogen-15 and carbon-13

abundances in forest soil organic matter. Soil Science Society of America

Journal 52(6), 1633-1640.

Ohsawa, M. (Ed.), 1987. Life Zone Ecology of the Bhutan Himalaya, Laboratory of

Ecology, Chiba Univerisity, Japan.

Peri, P.L., Ladd, B., Pepper, D.A., Bonser, S.P., Laffan, S.W., Amelung, W., 2012.

Carbon (δ13C) and nitrogen (δ15N) stable isotope composition in plant and soil in

Southern Patagonia's native forests. Global Change Biology 18(1), 311-321.

Prescott, C.E., 2002. The influence of the forest canopy on nutrient cycling. Tree

Physiology 22(15-16), 1193-1200.


Page | 105

Reich, P.B., Luo, Y., Bradford, J.B., Poorter, H., Perry, C.H., Oleksyn, J., 2014.

Temperature drives global patterns in forest biomass distribution in leaves,

stems, and roots. Proceedings of the National Academy of Sciences 111(38),

13721-13726.

Schaub, M., Alewell, C., 2009. Stable carbon isotopes as an indicator for soil

degradation in an alpine environment (Urseren Valley, Switzerland). Rapid

Communications in Mass Spectrometry 23(10), 1499-1507.

Schneider, C.A., Rasband, W.S., Eliceiri, K.W., 2012. NIH Image to ImageJ: 25 years of

image analysis. Nat Meth 9(7), 671-675.

Singh, S.K., Sarin, M.M., France-Lanord, C., 2005. Chemical erosion in the eastern

Himalaya: Major ion composition of the Brahmaputra and δ13C of dissolved

inorganic carbon. Geochimica et Cosmochimica Acta 69(14), 3573-3588.

Sutherland, R.A., van Kessel, C., Farrell, R.E., Pennock, D.J., 1993. Landscape-scale

variations in plant and soil nitrogen-15 natural abundance. Soil Science Society

of America Journal 57(1), 169-178.

Syed, F.S., Körnich, H., Tjernström, M., 2012. On the fog variability over south Asia.

Climate Dynamics 39(12), 2993-3005.

Tashi, S., Singh, B., Keitel, C., Adams, M., 2016. Soil carbon and nitrogen stocks in

forests along an altitudinal gradient in the eastern Himalayas and a meta-analysis

of global data. Global Change Biology 22(6), 2255-2268.

Torn, M.S., Trumbore, S.E., Chadwick, O.A., Vitousek, P.M., Hendricks, D.M., 1997.

Mineral control of soil organic carbon storage and turnover. Nature 389(6647),

170-173.


Page | 106

Vitousek, P., Field, C., Matson, P., 1990. Variation in foliar δ13C in Hawaiian

Metrosideros polymorpha: A case of internal resistance? Oecologia 84(3), 362-

370.

Wang, G., Zhou, L., Liu, M., Han, J., Guo, J., Faiia, A., Su, F., 2010. Altitudinal trends

of leaf δ13C follow different patterns across a mountainous terrain in north China

characterized by a temperate semi-humid climate. Rapid Communications in

Mass Spectrometry 24(11), 1557-1564.



broadleaved forest. Journal of Renewable Natural Resources of Bhutan 5(1), 55-

67.

Warren, C.R., McGrath, J.F., Adams, M.A., 2001. Water availability and carbon isotope

discrimination in conifers. Oecologia 127(4), 476-486.

Zhao, C., Chen, L., Ma, F., Yao, B., Liu, J., 2008. Altitudinal differences in the leaf

fitness of juvenile and mature alpine spruce trees (Picea crassifolia). Tree

Physiology 28(1), 133-141.

Zhou, Y., Fan, J., Zhong, H., Zhang, W., 2013. Relationships between altitudinal

gradient and plant carbon isotope composition of grassland communities on the

Qinghai-Tibet Plateau, China. Science China. Earth Sciences 56(2), 311-320.

Zimmermann, M., Leifeld, J., Conen, F., Bird, M., Meir, P., 2012. Can composition and

physical protection of soil organic matter explain soil respiration temperature

sensitivity? Biogeochemistry 107(1-3), 423-436.


Page | 107


Supplement 4.1 Leaf stomatal measurements for Persea sp. (Means ± SD).

Forest

Pore length (µm) Density (mm-2) gw max (molm-2 s-1)

Sunlit Shade Sunlit Shade Sunlit Shade

STF 10.0 ± 2.1 10.7 ± 1.7 908 ± 188 775 ± 160 4.1 ± 1.2 4.0 ± 1.3

WTBLF 9.8 ± 1.7 11.0 ± 1.5 869 ± 140 796 ± 206 4.0± 1.1 4.2 ± 1.2

CTBLF 11.4 ± 1.7 11.8 ± 1.5 855 ± 131 795 ± 132 4.8 ± 1.6 4.6 ± 0.8

Supplement 4.2 Leaf stomatal measurements for tree species from different forest

types

Pore length (µm) Density (mm-2) Gw max

TF 8.00a ± 1.33 688.37 a ± 294.82 3.15 a ± 1.74 STF 9.05 a ± 1.97 835.78 a ± 258.69 3.68 a ± 1.54 WTBLF 9.04 a ± 2.48 760.62 a ± 288.21 3.40 a ± 1.16 CTBLF 9.56 a ± 2.82 761.23 a ± 276.72 3.71 a ± 1.18 CTF 5.80 a ± 0.91 868.21 a ± 110.86 2.33 a ± 0.45

Supplement 4.3 Relationship of maximum potential stomatal conductance Gw max with

altitude


Page | 108

Supplement 4.4 Relationship of the difference of δ15N in the understorey leaf and soil

with altitude.

y = -0.0008x - 4.8778 r² = 0.1822

-12

-9

-6

-3

00 1000 2000 3000

(Δδ

15N

leaf

- so

il)

Altitude (m)


Page | 109

Supplementary Information Supplement 4.1 Pearson correlation coefficients between soil, plant and environmental parameters.

Alt Temp Soil N (%)

Soil C (%)

N (gkg-1)

C (gkg-1)

Soil δ15N

Soil δ13C

δ15N Lt

δ13CLt

δ15N leafU

δ13C leafU

δ15N leafC

δ13C leafC

δ13C woodC Clay Sand Silt C:N CEC EC

Alt 1

Temp -.93** 1

Soil N (%) .50* -.47* 1

Soil C (%) .66** -.61** .95** 1

Soil N (gkg-1) .50* -.47* 1.0** .95** 1

Soil C (gkg-1) .66** -.61** .95** 1.00*

* .95** 1.

Soil δ15N .31 -.23 .19 .29 .19 .29 1

Soil δ13C .08 -.02 -.30 -.23 -.30 -.23 .50* 1

δ15N Lt -.50* .53* -.12 -.27 -.12 -.27 .05 -.14 1

δ13C Lt .60** -.38 .14 .27 .14 .27 .58** .52* -.20 1

δ15N leafU -.51* .57** -.37 -.50* -.37 -.50* .05 .20 .86** -.08 1

δ13C leafU .25 -.17 .01 .06 .01 .06 .09 -.01 .14 .24 .15 1

δ15N leafC -.61** .58** -.12 -.25 -.12 -.25 -.11 .02 .53* -.42 .56** -.28 1

δ13C leafC -.14 .25 -.17 -.16 -.17 -.16 .30 .52* .14 .27 .25 -.27 .44 1

δ13C woodC .53* -.50* .12 .24 .12 .24 .49* .51* -.10 .48* .01 .07 -.05 .46* 1

Clay (%) .24 -.16 .33 .38 .33 .38 .28 .37 -.39 .27 -.32 .13 -.18 .00 .25

Sand (%) -.13 .17 -.31 -.30 -.31 -.30 -.09 .00 .43 .09 .38 .08 .09 .23 .11 -.65** 1

Silt (%) -.01 -.09 .14 .09 .14 .09 -.10 -.29 -.26 -.34 -.24 -.22 .03 -.30 -.34 .07 -.79** 1

C:N .83** -.78** .29 .53* .29 .53* .22 -.02 -.61** .37 -.62** .30 -.47* -.19 .35 .28 -.14 -.04 1

CEC .48* -.42 .65** .74** .65** .74** .24 .09 -.48* .22 -.57** .14 -.27 .02 .25 .72** -.53* .13 .51* 1

EC -.26 .39 -.01 -.09 -.01 -.09 -.27 -0.01 .05 .00 .07 -.32 .24 .43 -.15 .28 -.32 .20 -.35 .13 1

Carbon and nitrogen concentrations in soil are averages of 60 – 100 cm. Significant isotopes correlations are marked with * (< 0.05) and

** (< 0.01).

Chapter 5. Mineral-organic association

Page | 110

Chapter 5.

Mineral-organic associations and organic carbon forms in

forest soils at different altitudes of eastern Himalayas

Abstract

Soil organic carbon (SOC) stability and retention are influenced by soil mineralogy,

climate and land use land cover. However, biological and environmental processes interact

with different C forms and affect SOC storage and stability. To decipher these interactions,

we sequentially fractionated forest soils from an altitude transect between 317 and 3300 m in

Bhutan into different density fractions. X-ray diffraction was used to determined soil

mineralogy and diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy was

used to determine organic compounds and inorganic mineralogy of each soil density fraction.

Isotopic ratio mass spectroscopy was used to determine the C, nitrogen (N) concentrations

and δ13C for all soil density fractions. Carbon and N concentrations decreased with increasing

density fractions. The decreased C:N ratios, greater proportion of aromatic C and enrichment

of δ13C with increasing soil density fraction for all altitude forest soils suggests more

processed SOC associate with higher soil density fractions. Smaller Index 1, a metric of SOC

decomposition for highest altitude, cold temperate forest (CTF) and increasing C:N ratios for

all soil density fractions with increasing altitude suggests reduced decomposition at higher

altitude forests. In addition, Index 2, the relative recalcitrance for CTF at the highest altitude

is low, suggesting limited decomposition even for easily decomposable carboxyl groups and

polysaccharides. Therefore with advent of global warming, high altitude forest soil could

rapidly lose C through enhance decomposition further aggravating climate change.


Page | 111

5.1.1 Introduction

Soils are the largest terrestrial reservoir of organic carbon (OC), with an estimated

stock of between 3,500 and 4,800 Pg C in the top one meter worldwide (Lehmann and

Kleber, 2015). The global soil organic carbon (SOC) pool is 5 − 10 times larger than the

carbon (C) stocks in vegetation (Noble et al., 2000). Depending on land use and climatic

condition a variable amount of organic material is continually added to the soil. Biological,

physical and chemical processes in the soil transform organic material into organic fragments

and microbial products that are often intimately associated with mineral phases (Baisden et

al., 2002; Torn et al., 1997). The association of microbially– derived and other organic matter

(OM) with minerals is an important process that controls the long-term retention of OM in

soils (Baldock and Skjemstad, 2000; Lützow et al., 2006; Sollins et al., 2006; Torn et al.,

1997). Organic–mineral associations in soils can occur via physical interactions involving

weak adhesion to encapsulation and occlusion, and chemical bonding involving complexation

reactions (Lützow et al., 2006). These physical and chemical interactions between minerals

and organic matter decrease the bio-accessibility of OM, particularly in the case of

multilayered adsorption of organic compounds to mineral surfaces (Baldock and Skjemstad,

2000; Kleber et al., 2015). Therefore, the soil C contained in organic–mineral associations is

less prone to mineralisation and can remain in soil for an extended period of time. Indeed,

data based on radiocarbon studies show that turnover times for C present within organic–

mineral associations range between decades to millennia and are on average four times longer

than those of C in free or occluded OM (Kleber et al., 2015).

Density fractionation of soil allows the isolation of organic matter that is associated

with minerals from free organic matter or organic matter that has little interaction with

minerals. The procedure does not require any chemical treatment and relies on the difference

in density between minerals and organic materials with additional physical dispersion

(Christensen, 1992). Soil organic carbon in major Australian soil types under native

vegetation were found to separate into particulate organic matter, phyllosilicate dominant,

quartz and feldspars dominant and Fe oxide dominant fractions (Jones and Singh, 2014). The

surface interactions of each mineral pool influence change particular to chemical composition

of OM bonds and thereby associate with discrete organic functional groups, which in turn


Page | 112

determines the stability of the OM (Doetterl et al., 2015; Jones and Singh, 2014; Kleber et al.,

2007).

Apart from organo–mineral association, environmental and biological factors also

interact and affect SOC storage and stability. From a study on Changbai Mountain, China,

increasing temperatures were found to effect labile C pools and not stable C, whereas

decomposition of the intermediate pool was influenced by soil nitrogen availability (Tian et

al., 2016). However, environmental factors such as precipitation and temperature were found

to be of lesser significance compared to geochemical predictors for SOC (Doetterl et al.,

2015).

Diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy has been

effectively used to determine bands representing organic functional groups in soil organic

matter (SOM) (Ellerbrock and Gerke, 2013; Margenot et al., 2015; Veum et al., 2014;

Yeasmin et al., 2016). The main advantage of DRIFT spectroscopy is its capacity to

determine functional groups of SOM without the requirement for samples pretreatment

(Armaroli et al., 2004; McKenzie et al., 1984). Additionally obtaining DRIFT spectra of

samples is relatively a quick procedure.

With sequential density fractionation of bulk forest soils and DRIFT analysis we set

out i) to characterize C associated with different minerals in the SOM and ii) to determine the

proportion of C forms associated with different minerals for the different altitude forest soils.

The characterisation of C forms and the determination of minerals in the soils were carried

out to evaluate the stability of C associated with different soil minerals. Furthermore, the

proportion of C forms associated with different minerals in soils at different altitudes may be

useful to determine the proportion of C stocks associated with pools of varying turnover time.

5.2 Materials and Methods:

5.2.1 Study area

Soil samples were collected from five profiles at different locations along an altitude

gradient from 317 m to 3300 m in Bhutan. The composition of forest tree species changed

with increasing altitude. the location of each soil profile corresponded to a forest type based


Page | 113

on the classification proposed by Ohsawa (1987). The sites included - i) tropical forest (TF,

526 m a.s.l.), 26° 51'34.8" N, 89° 24' 10.6" E, ii) sub-tropical forest (STF, 1360 m a.s.l.), 26°

53'13.3" N, 89° 26' 52.1" E, iii) warm temperate broadleaf forest (WTBLF 1902 m a.s.l.), 26°

55' 17.4" N, 89° 29' 27.9" E, iv) cool temperate broadleaf forest (CTBLF 2551 m a.s.l.), 26°

56' 54.9" N, 89° 30' 17.7" E and v) cold temperate forest (CTF, 3098 m a.s.l.), 26° 59' 21.3"

N, 89° 31' 41.6" E.

Soil profiles were exposed to a depth of about 100 cm and soil samples (~1500 g)

were collected from the two top horizons. Soil samples were collected from the top two

genetic horizons to better evaluate the association of C with different soil minerals. Soil

samples were air dried, gently crushed by hand and then passed through a 2 mm sieve to

separate the fine soil fractions from the pebbles and stones. Soil pH was measured in a 1:5

soil-water suspension (Rayment and Higginson, 1992), and particle size analysis was

determined by the pipette method (Gee et al., 1986). Total C and N in the bulk soils were

analysed using a Thermo Finnigan Delta V isotope ratio mass spectrometer coupled to

ConfloIV and FlashHT peripherals (Thermo Fisher Scientific, Bremen, Germany). The total

C concentrations in the soil were considered as total organic C due to the absence of

carbonates in these highly acidic soils. General soil characteristics are presented in Table 5.1.

Table 5.1 Physico-chemical properties of soils from the two top genetic horizons of soil

profiles from forests at different altitudes in Bhutan.

Profile Forest Depth (cm)

pH Sand (%)

Silt (%)

Clay (% )

CEC (Cmolc kg-1)

Total C (%)

Total N (%)

C : N ratio altitude

(m) type (1:5 H2O)

526 Tropical 0−10 4.3 39 27 34 25.81 3.19 0.32 9.9 10–20 4.3 35 30 35 19.72 1.62 0.16 9.8

1360 Sub-tropical 0−25 4.3 35 42 24 37.24 6.50 0.56 11.6 25−53 4.5 31 37 32 31.89 3.94 0.30 13.0

1902 Warm temperate broad leaf

0−24 4.1 48 30 22 36.84 8.73 0.85 10.3

24−55 4.7 33 28 39 29.45 3.24 0.34 9.4 2551 Cool

temperate broad leaf

0−18 4.9 29 35 36 33.17 4.27 0.37 11.6

18−34 5.1 35 25 40 28.36 2.18 0.19 11.4 3098 Cold

temperate 0−17 4 34 39 27 37.27 8.08 0.54 14.9

17−32 4.4 34 39 27 37.18 7.03 0.41 16.9


Page | 114

5.2.2 Soil density fractionation

Homogenised soil samples collected from the two genetic horizons were used for

density fractionation procedure. Sodium polytungstate (SPT) was used to separate the soil

into four density fractions, i.e. < 1.8 g cm-3, 1.8 – 2.2 g cm-3, 2.2 – 2.6 g cm-3 and > 2.6 g cm-

3, following the procedure described by (Sollins et al., 2006; Yeasmin et al., 2016).

Approximately 30 g of air dried soil was placed into a 250 ml polycarbonate centrifuge bottle

and 125 ml of SPT solution of the lowest desired density (1.8 g cm-3) was added to the bottle,

vigorously shaken by hand and then placed on a shaker table at 300 rpm for 3 h to disperse

soil aggregates. For density fractions less than 1.8 g cm-3 suspensions were centrifuged at

~1100 g for 30 min (Spintron GT 175). The floating materials were aspirated with the SPT.

Glass fibre filter paper with pore size of 0.7 µm (Merek Millipore) was used to recover the

SPT which was returned to the same centrifuge bottle. The centrifuge bottle was shaken again

for 1 h on the shaker and centrifuged for 30 min at ~1100 g and floating material aspirated

with the SPT for second yield. Both yields for the same density fraction were combined and

thoroughly rinsed with deionised water to remove SPT residues. The removal of SPT residue

was confirmed by measuring the electrical conductivity (EC) of filtrate deionised water (< 50

µS cm-1). The other three density fractions (i.e. 1.8 – 2.2 g cm-3, 2.2 – 2.6 g cm-3 and > 2.6 g

cm-3) were separated following similar steps except the centrifugation was done at high speed

(Sorval RC-5C Super speed, ~23,000g). SPT density was successively adjusted for the next

higher density fraction by adding SPT and the procedure repeated to isolate the required

density fraction. Separated density fractions were oven dried at 40° C and then ground into a

fine powder using a mortar and pestle for further analyses. Only one sample from each of the

genetic horizons of the soil profiles were fractionated due to resource and time constraints.

5.2.3 Isotopic analysis

Total C, total N and δ13C in the bulk soils and density fractions were determined on

a Thermo Finnigan Delta V isotopic mass spectrometer. The precision for C and N

measurements were between 0.07 and 0.10%, and for δ13C was between 0.05 and 0.08‰.


Page | 115

5.2.4 Mineralogical analysis of soils

Clay minerals were identified by X-ray diffraction analysis of oriented and randomly

oriented specimens. The clay fraction (< 2 mm) was obtained by a dispersion−sedimentation

procedure (Klute, 1986). XRD patterns were obtained using monochromatic CuKα radiation

(35 kV and 28.5 mA) on a GBA MMA diffractometer (Diffraction Technology Pty. Ltd.

Australia). Phyllosilicates were identified from XRD analysis of the oriented samples after

standard pre-treatments (Brindley and Brown, 1980). Random powder XRD analysis of the

clay fractions and density fractions allowed the identification of all minerals in the samples.

5.2.5 Spectroscopic analysis of soil density fractions

DRIFT spectra of the bulk soils and density fractions of the soils from the two

horizons were obtained using a Bruker Vertex 80 V (Bruker Optics Germany) spectrometer

equipped with a Pike EasiDiffTM (Pike technologies , USA) DRIFT accessory. Soil samples,

density fractions and potassium bromide (KBr) were finely ground by hand with an agate

mortar and pestle. The ratio of soil or density fraction to KBr was kept to1:30 for the DRIFT

analysis.

Soil-KBr mixtures were kept in an oven at 40 °C overnight to minimize moisture

before the DRIFT analysis. Spectra were recorded in the 4000 − 600 cm-1 range at 4 cm-1

resolution and 64 scans were obtained for each sample. For each sample, three separate

replicates (sub-samples) were analysed. Spectral bands were assigned to organic compounds

and minerals based on the published literature (Table 5.2). Spectral processing, baseline

correction, and relative area peak calculations were done using GRAM /AITM software

(Version GRAM SUITE9.2, Thermo Fisher Scientific Inc.).

From the SOM functional groups identified, two indices were calculated to estimate

the proportions of decomposable and recalcitrant portion of OC in the soil density fractions.

Index 1 is the ratio of aromatic C to aliphatic C (Table 5.2) which has been hypothesized to

be a matrix of OC decomposition (Margenot et al., 2015). Index 2 is based on the ratio of C

to O functional groups (Table 5.2) and is considered to be a measure of the recalcitrance of

SOM (Margenot et al., 2015; Veum et al., 2014). We could not identify all bands of C forms

as used by Margenot et al. (2015) and Veum et al. (2014) to derive the two indices. Therefore,


Page | 116

the two indices were slightly modified and derived using the bands presented in the following

formulae:

Index 1 = 1650 cm‐1

2931cm‐1 + 2851cm‐1 Equation 5.1

Index 2 = 2931cm‐1 + 2851cm‐1 + 1650 cm‐1

1731cm‐1 + 1161cm‐1 Equation 5.2

Table 5.2 DRIFT spectra band assignment for organic and inorganic bands. ν, stretching

vibration; νas asymmetric stretching vibration; νs, symmetric stretching vibration and δ,

bending vibration.

Band region (cm-1) Assigned bands (cm-1)

Organics Inorganics

3710 – 3595 – O-H stretching of clay mineralsab

2960 – 2840 2931c aliphatic nas (C-H) and 2851c aliphatic ns (C-H)

2040 – 1750 – Quartza at 2000,1870 and 1790 cm–1

1740 – 1698 1731cd ns (C=O) carbonyl groups – 1660 – 1580 1650 aromatic n (C=C)c – 1560 – 1500 1540 δ(N-H) and n (C-N)d –

1170 – 1148 1163 nas C-O-Cc (Polysaccharides) – a (Nguyen et al., 1991); b (Saikia and Parthasarathy, 2010); c (Johnson et al., 2007); d (Mazurek et al.,

2013); (Johnston et al., 1996)

5.2.6 Statistical Analysis

IBM SPSS Statistics 21 (Armonk, NY, USA) was used to perform statistical

analyses. To compare variation for organic and inorganic bands in the DRIFT spectra for the

different density fractions across altitudinal gradient and soil depth, multivariate pairwise

comparisons were carried out. Homogenised soil samples were collected from a single soil

profile for each of the forest types. However, for all samples the DRIFT analysis was done

for three sub-samples; which were used to perform the pairwise comparison between surface

and subsurface soils. Correlations (Pearson product-moment correlation coefficients) between

organic and inorganic functional groups and soil properties were performed. Correlations

were considered significant at P < 0.05.


Page | 117

5.3 Results

5.3.1 Characterization of the different altitude forest soils

The mineral compositions of the different altitude forest soils were similar and

mostly composed of chlorite, mica, interstratified 2:1 clay minerals, kaolinite, quartz and

feldspars (Fig. 5.1 and 5.2). The lowest altitude (TF) and highest altitude (CTF) soils were

dominated by mica in both the surface and sub-surface soils. Chlorite featured prominently in

the surface and sub-surface soils of STF and kaolinite in the mid-altitude (WTBLF, Fig. 5.2).

Interstratified 2:1 clay minerals featured in all surface and sub-surface soils. With increasing

soil density, the proportion of quartz increased for all altitude forest soils Phyllosilicate

contents were highest in the two lowest density fractions (< 1.8 and 1.8 − 2.2 g cm-1) and

lowest for the highest density fraction (Supplement 5.1).

Fig. 5.1 Random powder X ray diffraction patterns of different altitude forest soils a) surface

soils b) sub-surface soils. TF = tropical forest, STF = sub-tropical forests, WTBLF = warm

temperate broadleaf forest, CTBLF = Cool temperate broadleaf forest and CTF = cold

temperate forests Ch: Chlorite, M: mica (illite), K: kaolinite, Q: quartz, Fld: feldspars


Page | 118

Fig. 5.2 X-ray diffraction patterns of the oriented clay fractions of different altitude forest

soils, patterns from (a) to (e) are for the surface soils and patterns from (f) to (j) are for sub-

surface soils. TF = tropical forest, STF = sub-tropical forests, WTBLF = warm temperate

broadleaf forest, CTBLF = cool temperate broadleaf forest and CTF = cold temperate forests

Ch: Chlorite, M: mica (illite), IS: Interstratified 2:1 clay minerals; K: kaolinite, Q: quartz,

Fld: feldspars


Page | 119

5.3.2 Properties for soil density fractions

The density fraction less than 1.8 g cm-3 comprised predominantly particulate

organic matter (POM), while the density fractions greater than 1.8 g cm-3 consisted of organic

matter associated with minerals (1.8 − 2.2 g cm-3, 2.2 − 2.6 g cm-3 and > 2. 6 g cm-3).

The proportion of POM in the total SOM was consistently greater in the surface

horizons compared to the sub-surface horizons. The particulate organic matter proportion in

the surface horizon soil samples increased from 1.3% in the lowest altitude soil (TF) to

45.8% (Fig. 5.3) for the highest altitude soil (CTF). Similarly, the proportion of POM in the

total SOM in the sub-surface horizon increased from 0.75% in the lowest altitude forest soils

to 5.2% in the highest altitude forest soil. In contrast, the proportion of the heaviest density

fraction (> 2.6 g cm-3) was consistently greater for the sub-surface genetic soil horizon

compared to the surface soil horizon, with the exception for CTBLF. The proportion of the

heaviest density fraction (>2.6 g cm-3) decreased with altitude for the sub-surface soil from

85.8% to 42.3% and once again the results for the CTBLF sample were inconsistent with the

trend. The 2.2 − 2.6 g cm-3 density fraction constituted the dominant fraction from most soils,

except for the CTF surface soil where the POM was the dominant fraction; the TF sub-

surface soil where > 2.6 g cm-3 was the dominant fraction and in WTBLF surface soil where

1.8 − 2.2 g cm-3 fraction was most dominant.

The proportion of surface soil C and N content in the fractions < 2.2 g cm-3 generally

increased with increasing altitude, with the exception in the CTBLF. In contrast, in the > 2.2

g cm-3 fractions the C and N proportion decreased with increasing altitude, except for CTBLF

(Fig. 5.3a−e). Accordingly the lowest altitude surface and sub-surface soils (Fig. 5.3a&f)

stored a greater proportion of the C and N in the heaviest density fraction. In contrast the

lower density soil fractions at highest altitude (CTF) stored the greatest proportion of the C

and N.


Page | 120

Fig. 5.3 Proportion of soil mass, total C and total N distribution in density fractions of (a − e)

surface and (f − j) sub-surface soils for various altitude forest (a & f) TF, (b & g) STF, (c &

h) WTBLF, (d & i) CTBLF, (e & j) CTF, TF = tropical forest, STF = sub-tropical forests,

WTBLF = warm temperate broadleaf forest, CTBLF = cool temperate broadleaf forest and

CTF = cold temperate forests.


Page | 121

Carbon (15.3 − 29.2%) and N (0.91 − 1.89%) concentrations were highest in the

POM fraction of the soils and both concentrations decreased with increasing density of the

fractions in all soils (Fig. 5.4). The POM fraction of the STF from both surface and sub-

surface horizons contained the highest C concentration (29.2% and 28.0%, Fig. 5.4a&b). The

heaviest density fraction returned the lowest C and N concentrations amongst the density

fractions of all soils. Carbon and N concentrations in the fractions decreased with increasing

density and the concentrations were generally greater for higher altitude soils within a

particular density fraction.

C:N ratio was greatest in the POM fractions (16.2 to 30.7) and decreased with

increasing density of soil fractions irrespective of the altitude (Fig. 5.4e&f). The C:N ratio in

the POM of soils from the sub-surface horizon was consistently greater than the

corresponding ratio in the surface horizon samples. The heaviest density fraction for each soil

had the lowest C:N ratio which ranged between 8.1 and 15.4. δ13C became enriched with soil

depth and with increasing density fractions, irrespective of altitude (Fig. 5.4g&h). δ13C values

for the surface soils increased with altitude of the forest irrespective of density fractions.


Page | 122

Fig. 5.4 Total soil (a & b) C and (c & d) N concentrations, (e & f) C:N ratios and (g & h)

δ13C values in different density fractions for various altitude forest soils. TF = tropical forest,

STF = sub-tropical forests, WTBLF = warm temperate broadleaf forest, CTBLF = cool

temperate broadleaf forest and CTF = cold temperate forests


Page | 123

5.3.3 DRIFT analysis

DRIFT spectra for different density fractions of various altitude forest soils are

presented in Fig. 5.5a − e. Two distinct bands at 3697 and 3620 cm-1 were observed in the

spectra of all density fractions and the bulk soils, except the lowest altitude soil fraction,

where only the 3620 cm-1 band was noticeable. The bands at 3697 and 3620 cm-1 were

assigned to OH stretching of kaolinite and other phyllosilicates (Nguyen et al., 1991). XRD

analysis of the density fractions and the clay fractions of soils showed the presence of

kaolinite, chlorite, mica and an interstratified 2:1 clay mineral (Supplement 5.1 and Fig. 5.2).

A broad band centred at 3400 cm-1 (Fig. 5.5), in the spectra for all density fractions was

attributed to ν (OH) of sorbed water, hydrous minerals and ν (NH) (Singh et al., 2016).

Bands due to aliphatic ν (C−H) stretching were observed at 2931 and 2851 cm–1

(Johnson et al., 2007). Relative intensities of these bands were greater in lower density

fractions (< 2.2 g cm-3) and higher altitude forest soils (3098 m). The bands at 2000, 1870,

1790 cm–1 were attributed to quartz (Nguyen et al., 1991). Quartz featured prominently in the

random powder XRD patterns of different density fractions of all soils (Supplement 5.1). The

relative proportion of quartz was greater in the higher density fractions and higher altitude

soils. High altitude soils showed a band at 1731 cm–1 that was assigned to the carboxyl

groups of SOM (Johnson et al., 2007; Mazurek et al., 2013). A distinct bands was present at

1650 cm–1 in all soil density fractions and this was attributed to aromatic ν (C = C) (Johnson

et al., 2007). The band at 1540 cm–1 was assigned to amides (Mazurek et al., 2013) and at

1163 cm–1 to polysaccharides (Johnson et al., 2007).


Page | 124

Fig. 5.5 DRIFT spectra of the four density fractions (< 1.8, 1.8−2.2, 2.2− 2.6 and > 2.6) of

surface soils from (a) TF = tropical forest (b) STF = sub-tropical forest (c) WTBLF = warm

temperate broadleaf forest (d) CTBLF = cool temperate broadleaf forest and (e) CTF = cold

temperate forest. Bands for kaolinite and other phyllosilicates at 3697 and 3620 cm-1 (Nguyen

et al., 1991; Saikia and Parthasarathy, 2010), for aliphatic C at 2931 and 2851 cm-1 (Johnson

et al., 2007; Veum et al., 2014), for quartz at 2000,1870 and 1790 cm-1 (Nguyen et al., 1991),

characteristic carboxyl C band at 1731 cm-1 (Johnson et al., 2007; Mazurek et al., 2013),

aromatic band at 1648 cm-1 and 1610 cm-1 (Baes and Bloom, 1989; Cocozza et al., 2003;

Veum et al., 2014), amide band at 1540 cm-1 (Mazurek et al., 2013) and polysaccharides band

at 1161 cm-1 (Demyan et al., 2012; Johnson et al., 2007).


Page | 125

The proportion of aliphatic C decreased with soil depth, except in the highest

altitude forest soil where it increased slightly. The proportion of aliphatic C decreased with

increasing density for both surface (Fig. 5.6a) and sub-surface (Fig. 5.6b) soils irrespective of

the altitude. Higher altitude soils generally contained a greater proportion of aliphatic C

across all density fractions with a few exceptions.

In contrast to aliphatic C, the proportion of aromatic C in the density fractions increased with

increasing density for all soils (Fig. 5.6c&d). The CTF soil fractions contained much smaller

proportion of aromatic C as compared to the other four soils. The carboxyl group was greater

in the POM fraction than the mineral-associated heavy fractions of all soils (Fig. 5.6e&f). In

contrast, polysaccharides proportion was consistently greater in the highest density fraction

than the other fractions of all soils (Fig. 5.6g&h). The proportion of amides was greater in the

surface horizons than in the corresponding sub-surface horizons across all density fractions

(Fig. 5.6i&j). Within surface and sub-surface horizons, the proportion of amide was generally

similar across different soil density fractions, except for the STF surface soil and WTBLF

sub-surface soil, where it is was much greater in the 2.2 – 2.6 density fraction.


Page | 126

.

Fig. 5.6 Relative integrated peak area of organic bands in the DRIFT spectra density fractions

(a & b) aliphatic C, (c & d) aromatic C, (e & f) carboxyl, (g & h) polysaccharides, and (i & j)

amides in surface and sub-surface soils. The samples have been separated based on the forest

types i.e. TF = tropical forest, STF = sub-tropical forests, WTBLF = warm temperate

broadleaf forest, CTBLF = cool temperate broadleaf forest and CTF = cold temperate forests.


Page | 127

Index 1 which is based on the ratio of aromatic C to aliphatic C and is a matrix of

decomposition, increased with increasing density (Fig. 5.7 a&b). Sub-surface samples (Fig.

5.7b) had a larger value as compared to surface samples (Fig. 5.7a); except for the highest

altitude soil (CTF). Across the altitudinal gradient, samples from the highest altitude (CTF)

soil had the lowest Index 1 value for a corresponding density fraction as compared to the

other soils.

Index 2, which is a measure of recalcitrant C, was greatest in the 1.8 − 2.2 g cm-3

fraction of all soils (Fig. 5.7c&d). Across different altitude forest soils, Index 2 value was

lower for the three fractions < 2.6 g cm-3 of the highest altitude soil (CTF). Surface soils had

lower recalcitrant C compared to the sub-surface soil at all altitudes, except for the heaviest

density fraction (> 2.6 g cm-3) where Index 2 values were similar between various altitude

soils.

Fig. 5.7 Indices calculated from the relative integrated peak area of organic bands in the

DRIFT spectra of density fractions. Index 1 the matrix of decomposition for (a) surface soil,

(b) sub-surface soils and Index 2 a measure for C recalcitrance for (c) surface and (d) sub-

surface soils from different altitudes forests. TF = tropical forest, STF = sub-tropical forests,

WTBLF = warm temperate broadleaf forest, CTBLF = cool temperate broadleaf forest and

CTF = cold temperate forests.


Page | 128

Aliphatic C in the soils was positively correlated (p < 0.01) and aromatic C was

negatively correlated (p < 0.001) with altitude (Table 5.3). Total C and N concentrations, and

C:N ratio in the soils were positively correlated with aliphatic C (C = p < 0.001, N = p <

0.001 and C:N at p < 0.001) and negatively correlated with aromatic C (C = p < 0.001, N = p

< 0.0001 and C:N at p < 0.001). Aromatic C and Index 1 were correlated with soil depth (p =

<0.01). δ13C was positively correlated with aromatic C (p < 0.001) and negatively correlated

with aliphatic C (p < 0.001) Clay content was positively correlated with aromatic C (p <

0.01) and Index 1 (p = 0.05), and negatively with aliphatic C (p = 0.019). In contrast, silt

content was negatively correlated with aromatic C (p = 0.035) and positively correlated with

aliphatic C (p = 0.025).


Page | 129

Table 5.3 Pearson correlations for DRIFT bands representing organic and inorganic functional groups and soil properties

Altitude Depth Aliphatic

C Aromatic

C Index

1 Index

2 DF C (%) N

(%) δ13C C:N Clay

(%) Silt (%)

Sand (%)

Altitude 1 Depth .00 1.00 Aliphatic C .24** -.13 1 Aromatic C -.32** .34** -.93** 1 Index 1 -.02 .25** -.80** .77** 1 Index 2 -.24** .21* .30** -.07 -.30** 1 DF .0 .00 -.73** .57** .54** -.47** 1 C (%) -.03 -.04 .75** -.64** -.55** .33** -.92** 1 N (%) -.10 -.22 .72** -.65** -.55** .44** -.88** .90** 1 δ13C .41** .46** -.50** .49** .54** -.14 .59** -.69** -.71** 1 C:N ratio .19 .22 .68** -.54** -.43** .08 -.70** .80** .50** -.43** 1 Clay (%) -.08 .51 -.72* .79** .80** .12 - -.45 -.35 .60 -.19 1 Silt (%) .28 -.27 .70* -.67* -.71* .14 - .74* .54 -.29 .41 -.66* 1 Sand (%) -.23 -.33 .09 -.21 -.18 -.31 - -.29 -.18 -.41 -.24 -.49 -.33 1

Correlations coefficients are considered significant at * 0.05 and** 0.01 level. DF= density fraction


Page | 130

5.4 Discussion

5.4.1 Relationship between soil properties, soil density fractions and various

altitude soils

Particulate organic matter at all altitudes soil had the highest C and N

concentrations (Fig. 5.4). This is most likely due to relatively large amounts of plant

materials contained in POM (Baisden et al., 2002; Golchin et al., 1994). Mineral

associated or heavier soil density fractions (> 1.8 g cm-3) had lower C and N

concentrations. The lower altitude soils generally had a large proportion of heavier

density soil fractions > 1.8 g cm-3, which resulted in relatively greater proportion of

mineral associated C stored in these soils. The heavier density fractions usually have

a longer turnover rate than the POM in the lower density fractions (Crow et al.,

2007), which is heightened at lower altitude due to greater temperatures. In contrast,

the highest altitude surface (CTF) soils had a greatest proportion of the lightest soil

density fraction or POM and also a greatest proportion of C stored in the POM

compared to the other soils. These high altitude forests had greater input of OM

(Tashi et al., 2016) and also usually accumulate SOM in POM due to slower

decomposition rate because of lower temperatures (Conant et al., 2011; Davidson

and Janssens, 2006). The mineral associated fractions at the lower altitudes were

relatively more δ13C enriched and had lower C:N ratio compared to the

corresponding density fractions in the higher altitude soils. Similarly with increasing

soil density fractions, δ13C became more enriched and C:N ratio became lower. This

implies microbially more processed SOC was associated with lower altitude soils and

with higher density of soil fractions, with strong mineral−organic interactions

(Baisden et al., 2002; Jones and Singh, 2014; Nadelhoffer and Fry, 1988).

Consequently C present in the mineral associated fraction is likely to be more stable

(Conen et al., 2008). In contrast, high C:N ratios in the POM contained in the lightest

soil density fraction (Fig. 5.4e&f) is probably due to the high content of fresh plant

residues that are at an early stage of decomposition (Golchin et al., 1994). These

results suggest that C in the POM (< 1.8 g cm -3) of soils have faster turnover rates

than the mineral associated C in the heavier soil fractions (Six et al., 1998).

Consequently C concentrations are negatively correlated with density (Table 5.3) and


Page | 131

similar results have been reported elsewhere (Sollins et al., 2006). Additionally,

increasing proportion of carboxyl C with decreasing density and increasing altitude

of the soils suggests that more labile C was present in the lighter soil density

fractions, especially in the high altitude forest soils. Furthermore the consistent δ13C

enrichment with increasing density suggests that more microbially processed C

(Nadelhoffer and Fry, 1988) is stored in the greater density fractions.

Lower C:N ratios across all soil density fractions with decreasing altitude

suggests greater mineralisation of OM that occurs with increasing temperature

(Garten Jr and Hanson, 2006; Tashi et al., 2016). Additionally, the decreasing

gradient of change of δ13 C with increasing altitude of the fractions (Fig. 5.4g&h)

indicates that SOC mineralisation decreased with decreasing temperature at high

altitudes. Thus, more C is expected to accumulate at higher altitudes (Garten Jr and

Hanson, 2006; Trumbore, 1993).

The positive correlation of aliphatic C with altitude and negative correlation

with δ13C indicates that high altitude forest soils accumulate greater amounts of

aliphatic C, mainly due to limited decomposition. Additionally, the negative

correlation of aromatic C with altitude and C:N ratio implies that lower altitude

forest soils store a greater proportion of the microbially processed C than the higher

altitude forest. Accordingly, turnover time for organic matter have been reported to

increase with increasing density fraction (Conen et al., 2008; Sollins et al., 2006).

5.4.2 Organo-mineral association

The proportion of aromatic C increased with increasing density of soil

fractions and with soil depth, except for the highest altitude (CTF) soil.

Consequently, ratio of aromatic C to aliphatic C was greater in heavier density

fractions and in the sub-surface soils, which indicates SOC was more decomposed or

microbially processed in the sub-surface soils compared to the surface soils (Baisden

et al., 2002). With progressive SOM decomposition, aliphatic C (2930 cm-1 and 2850

cm-1) decreases and aromatic C (1650 cm-1) increases relative to each other (Chefetz

et al., 1998). Index 1, which is a matrix of SOM decomposition, had significantly

lower values for the high altitude (CTF) soil fractions than the equivalent fractions of


Page | 132

soils at lower altitude. These results indicate a reduced decomposition rate of SOM at

high altitude, typically restricted by low temperatures (Mueller et al., 2015; Tashi et

al., 2016). The C:N ratios of the higher altitude soils were relatively larger in all soil

density fractions (Fig. 5.4e&f) that supports the hypothesis of decreased rate of SOM

decomposition (Baisden et al., 2002). Accordingly, there was a greater proportion of

aliphatic C at higher altitude and similar or reduced proportion of aromatic C across

forest types. Increased temperatures were reported to exert greater influence on

aliphatic C than on aromatic C (Tian et al., 2016). Zeraatpishe and Khormali (2012)

from their study in Golestan province in Iran, reported that climatic factors had a

greater influence on the SOC stock than clay minerals. While other studies have

reported that organo-mineral interactions are responsible for majority of the stable

SOM (Hassink, 1997; Mikutta et al., 2006).

The X-ray diffraction analysis of the highest altitude POM (Fig. 5.2 and

Supplement 5.1) suggests low levels of phyllosilicates. However, the CEC (Table

5.1) values for the high altitude soils are relatively high, which could be due to the

large OM (Helling et al., 1964) content in the high altitude soils. The POM fraction

in the highest altitude soil constitutes a large proportion (63%) of the aliphatic C .

Aliphatic C is a more active C pool and more easily metabolized (Demyan et al.,

2012). In contrast, the X-ray diffraction of low altitude (TF & STF) POM indicates

an abundance of clay minerals and a greater proportion of aromatic C in the

corresponding density compared to higher altitude soils. In general the X ray

diffraction of the different density fractions showed an increasing trend in the

relative proportion of clay minerals; particularly mica and chlorite, with increasing

density fractions (Supplement 5.1). Relative content of quartz and feldspars also

increased with increasing density fractions. This corresponded well with the

increased proportion of aromatic C and Index 1 with increasing density fractions.

Index 2 (ratio of C- to O- functional group) is a measure of relative

recalcitrance of SOM. Index 2 for lighter density fractions of the sub-surface soils

(Fig. 7d) was greater for lower altitude soils. As decomposition of organic matter

progresses, oxygen containing compounds are preferentially used by microbes and

therefore leaving behind more recalcitrant forms of C (Chefetz et al., 1998; Veum et


Page | 133

al., 2014). In contrast CTF at the highest altitude, even with high proportion of C, the

recalcitrance Index 2 is low, suggesting a limited decomposition even for easily

decomposable carboxyl and polysaccharides. The similar δ13C values across all soil

density fractions for CTF (Fig. 5.4g&h) support the argument that decomposition

was limited in the highest altitude forest soils. Our results show that there was large

amount of OC in high altitude soils; however, the C was largely in easily

decomposable C forms that may be more vulnerable decomposition at increased

temperatures with global warming. Mountainous regions makes up one fifth of the

global area and even small losses of SOC from this large C pool could have

significant impact on the atmospheric CO2 which may further aggravate climate

change.


Page | 134

References

Armaroli, T., Bécue, T., Gautier, S., 2004. Diffuse reflection infrared spectroscopy

(DRIFTS): Application to the in situ analysis of catalysts. Oil & Gas Science

and Technology 59(2), 215-237.

Baes, A., Bloom, P., 1989. Diffuse reflectance and transmission Fourier transform

infrared (DRIFT) spectroscopy of humic and fulvic acids. Soil Science




Global Biogeochemical Cycles 16(4),1-16.

Baldock, J.A., Skjemstad, J.O., 2000. Role of the soil matrix and minerals in

protecting natural organic materials against biological attack. Organic

Geochemistry 31(7–8), 697-710.

Brindley, G.W., Brown, G., 1980. Crystal Structures of Clay Minerals and their X-

Ray Identification, 5. Mineralogical Society of Great Britain and Ireland.

Chefetz, B., Adani, F., Genevini, P., Tambone, F., Hadar, Y., Chen, Y., 1998.

Humic-acid transformation during composting of municipal solid waste.

Journal of Environmental Quality 27(4), 794-800.

Christensen, B.T., 1992. Physical fractionation of soil and organic matter in primary

particle size and density separates, Advances in Soil Science. Springer, pp. 1-

90.

Cocozza, C., D'Orazio, V., Miano, T.M., Shotyk, W., 2003. Characterization of solid

and aqueous phases of a peat bog profile using molecular fluorescence

spectroscopy, ESR and FT-IR, and comparison with physical properties.

Organic Geochemistry 34(1), 49-60.

Conant, R.T., Ryan, M.G., Ågren, G.I., Birge, H.E., Davidson, E.A., Eliasson, P.E.,

Evans, S.E., Frey, S.D., Giardina, C.P., Hopkins, F.M., 2011. Temperature


Page | 135

and soil organic matter decomposition rates–synthesis of current knowledge

and a way forward. Global Change Biology 17(11), 3392-3404.

Conen, F., Zimmermann, M., Leifeld, J., Seth, B., Alewell, C., 2008. Relative

stability of soil carbon revealed by shifts in δ15N and C:N ratio.

Biogeosciences 5(1), 123-128.

Crow, S.E., Swanston, C.W., Lajtha, K., Brooks, J.R., Keirstead, H., 2007. Density

fractionation of forest soils: Methodological questions and interpretation of

incubation results and turnover time in an ecosystem context.


Davidson, E.A., Janssens, I.A., 2006. Temperature sensitivity of soil carbon

decomposition and feedbacks to climate change. Nature 440(7081), 165-173.

Demyan, M., Rasche, F., Schulz, E., Breulmann, M., Müller, T., Cadisch, G., 2012.

Use of specific peaks obtained by diffuse reflectance Fourier transform mid‐

infrared spectroscopy to study the composition of organic matter in a Haplic

Chernozem. European Journal of Soil Science 63(2), 189-199.

Doetterl, S., Stevens, A., Six, J., Merckx, R., Van Oost, K., Pinto, M.C., Casanova-

Katny, A., Muñoz, C., Boudin, M., Venegas, E.Z., 2015. Soil carbon storage

controlled by interactions between geochemistry and climate. Nature

Geoscience 8(10), 780-783.

Ellerbrock, R.H., Gerke, H.H., 2013. Characterization of organic matter composition

of soil and flow path surfaces based on physicochemical principles—A

review. Advances in Agronomy 121, 117-177.



Gee, G.W., Bauder, J.W., Klute, A., 1986. Particle-size analysis. Methods of Soil

Analysis. Part 1. Physical and mineralogical methods. Soil Science Society of

America, Madison


Page | 136

Golchin, A., Oades, J., Skjemstad, J., Clarke, P., 1994. Soil structure and carbon

cycling. Soil Research 32(5), 1043-1068.

Hassink, J., 1997. The capacity of soils to preserve organic C and N by their

association with clay and silt particles. Plant and Soil 191(1), 77-87.

Helling, C.S., Chesters, G., Corey, R.B., 1964. Contribution of Organic Matter and

Clay to Soil Cation-Exchange Capacity as Affected by the pH of the

Saturating Solution1. Soil Science Society of America Journal 28(4), 517-

520.

Johnson, E.J., Dorot, O., Liu, J., Chefetz, B., Xing, B., 2007. Spectroscopic

characterization of aliphatic moieties in four plant cuticles. Communications

in Soil Science and Plant Analysis 38(17-18), 2461-2478.

Johnston, C., Aochi, Y., Sparks, D., Page, A., Helmke, P., Loeppert, R., Soltanpour,

P., Tabatabai, M., Sumner, M., 1996. Fourier transform infrared and Raman

spectroscopy. Methods of soil analysis. Part 3-chemical methods., 269-321.

Jones, E., Singh, B., 2014. Organo-mineral interactions in contrasting soils under

natural vegetation. Frontiers in Environmental Science 2, 1-15.

Kleber, M., Eusterhues, K., Keiluweit, M., Mikutta, C., Mikutta, R., Nico, P.S.,

2015. Mineral–organic associations: Formation, properties, and relevance in

soil environments, Advances in Agronomy, pp. 1-140.

Kleber, M., Sollins, P., Sutton, R., 2007. A conceptual model of organo-mineral

interactions in soils: self-assembly of organic molecular fragments into zonal

structures on mineral surfaces. Biogeochemistry 85(1), 9-24.

Klute, A., 1986. Methods of soil analysis. Part 1. Physical and mineralogical

methods. American Society of Agronomy, Inc. Monographs 9(1), Madison,

Wisconsin, 1188 pp


Page | 137

Lehmann, J., Kleber, M., 2015. The contentious nature of soil organic matter. Nature

528(7580), 60-68.

Lützow, M.V., Kögel‐Knabner, I., Ekschmitt, K., Matzner, E., Guggenberger, G.,

Marschner, B., Flessa, H., 2006. Stabilization of organic matter in temperate

soils: mechanisms and their relevance under different soil conditions–a

review. European Journal of Soil Science 57(4), 426-445.

Margenot, A.J., Calderón, F.J., Bowles, T.M., Parikh, S.J., Jackson, L.E., 2015. Soil

organic matter functional group composition in relation to organic carbon,

nitrogen, and phosphorus fractions in organically managed tomato fields. Soil

Science Society of America Journal 79(3), 772-782.

Mazurek, S., Mucciolo, A., Humbel, B.M., Nawrath, C., 2013. Transmission Fourier

transform infrared microspectroscopy allows simultaneous assessment of

cutin and cell-wall polysaccharides of Arabidopsis petals. The Plant Journal

74(5), 880-891.

McKenzie, M., Culler, S., Koenig, J., 1984. Applications of diffuse reflectance FT-

IR to the characterization of an E-glass fiber/γ-APS coupling agent system.

Applied Spectroscopy 38(6), 786-790.

Mikutta, R., Kleber, M., Torn, M.S., Jahn, R., 2006. Stabilization of soil organic

matter: association with minerals or chemical recalcitrance? Biogeochemistry

77(1), 25-56.

Mueller, K.E., Hobbie, S.E., Chorover, J., Reich, P.B., Eisenhauer, N., Castellano,

M.J., Chadwick, O.A., Dobies, T., Hale, C.M., Jagodziński, A.M., Kałucka,

I., Kieliszewska-Rokicka, B., Modrzyński, J., Rożen, A., Skorupski, M.,

Sobczyk, Ł., Stasińska, M., Trocha, L.K., Weiner, J., Wierzbicka, A.,

Oleksyn, J., 2015. Effects of litter traits, soil biota, and soil chemistry on soil

carbon stocks at a common garden with 14 tree species. Biogeochemistry

123(3), 313-327.


Page | 138



Journal 52(6), 1633-1640.

Nguyen, T., Janik, L., Raupach, M., 1991. Diffuse reflectance infrared fourier

transform (DRIFT) spectroscopy in soil studies. Soil Research 29(1), 49-67.

Noble, I., Bolin, B., Ravindranath, N., Verardo, D., Dokken, D., (Ed), 2000. Land

Use, Land‐Use Change, and Forestry. Cambridge University Press.



Rayment, G., Higginson, F.R., 1992. Australian Laboratory Handbook of Soil and

Water chemical methods. Inkata Press Pty Ltd, Melbourne, Australia.

Saikia, B.J., Parthasarathy, G., 2010. Fourier transform infrared spectroscopic

characterization of kaolinite from Assam and Meghalaya, Northeastern India.

Journal of Modern Physics 1(04), 206.

Singh, B., Fang, Y., Johnston, C.T., 2016. A Fourier-Transform Infrared Study of

Biochar Aging in Soils. Soil Science Society of America Journal 80(3), 613-

622.

Six, J., Elliott, E., Paustian, K., Doran, J., 1998. Aggregation and soil organic matter

accumulation in cultivated and native grassland soils. Soil Science Society of


Sollins, P., Swanston, C., Kleber, M., Filley, T., Kramer, M., Crow, S., Caldwell,

B.A., Lajtha, K., Bowden, R., 2006. Organic C and N stabilization in a forest

soil: Evidence from sequential density fractionation. Soil Biology and

Biochemistry 38(11), 3313-3324.


forests along an altitudinal gradient in the eastern Himalayas and a meta-

analysis of global data. Global Change Biology 22(6), 2255-2268.


Page | 139

Tian, Q., He, H., Cheng, W., Bai, Z., Wang, Y., Zhang, X., 2016. Factors controlling

soil organic carbon stability along a temperate forest altitudinal gradient.

Scientific Reports 6, 18783.

Torn, M.S., Trumbore, S.E., Chadwick, O.A., Vitousek, P.M., Hendricks, D.M.,

1997. Mineral control of soil organic carbon storage and turnover. Nature

389(6647), 170-173.

Trumbore, S.E., 1993. Comparison of carbon dynamics in tropical and temperate

soils using radiocarbon measurements. Global Biogeochemical Cycles 7(2).

Veum, K.S., Goyne, K.W., Kremer, R.J., Miles, R.J., Sudduth, K.A., 2014.

Biological indicators of soil quality and soil organic matter characteristics in

an agricultural management continuum. Biogeochemistry 117(1), 81-99.

Yeasmin, S., Singh, B., Johnston, C.T., Sparks, D.L., 2016. Evaluation of pre-

treatment procedures for improved interpretation of mid infrared spectra of

soil organic matter. Geoderma.

http://dx.doi.org/10.1016/j.geoderma.2016.04.008

Zeraatpishe, M., Khormali, F., 2012. Carbon stock and mineral factors controlling

soil organic carbon in a climatic gradient, Golestan province. Journal of Soil

Science and Plant Nutrition 12(4), 637-654.

http://dx.doi.org/10.1016/j.geoderma.2016.04.008


Page | 140


Supplement 5.1 Random powder X ray diffraction patterns of different density

fractions of soils from a) Tropical forest b) Sub tropical forest, c) Warm temperate

broadleaf forest, d) Cool temperate broadleaf forest and e) Cold temperate forest. Ch:

Chlorite, M: mica (illite), K: kaolinite, Q: quartz, Fld: feldspars

Chapter 6. Allometric biomass equations

Page | 141

Chapter 6.

Allometric equation for biomass and carbon stocks of

forest along an altitudinal gradient in the eastern

Himalayas3

Abstract

Allometric equations remain essential to the estimation of forest biomass

and carbon (C) stocks. Most equations have been developed for American and

European forests, and transferring them to other species or eco-regions is

problematic. We lack all but the most rudimentary biomass equations for the large

forested areas in the eastern Himalayas, and this hinders reliable estimates of C

stocks. We destructively harvested 144 trees with diameters ranging from 10 cm to

77 cm, from five different forest types along an altitudinal gradient from 317 to 3300

m a.s.l. and used these to construct allometric equations. Model selection was based

on the Akaike Information Criterion (AIC), root mean square error (RMSE),

coefficient of determination (r2) of regressions, and absolute average deviation from

the measured aboveground biomass (AGB). Out of six models, we identified two that

could predict AGB across a range of trees, including those at the upper and the lower

ends of the scale. The models were mainly a function of diameter at breast height

(DBH) (log AGB = α + β log (DBH2), and with height (H) as an additional factor for

lower altitude forest types (log AGB = α + β log (DBH2 × H)). Inclusion of specific

gravity (SG) of wood improved models for three of the five forest types. We provide

both types of models and argue that wood SG should be collected during forest

3 This chapter has been accepted for publication in Forestry under the title “Allometric equation for biomass and carbon stocks of forest along an altitudinal gradient in the eastern Himalayas” in January 2017. Authors are Sonam Tashi, Claudia Keitel, Balwant Singh and Mark Adams


Page | 142

inventories as it is an important predictive variable especially for mixed species

forests. Average deviation of measured and estimated AGB ranged between 15.9 to

38.5% for the various models for the different forest types. With the best-fit models,

estimated aboveground C stocks increased with altitude from 57 to 207 Mg C ha-1.

The use of measured C concentration rather than an assumption of 50% of biomass

reduced estimated AGB C stocks between 6.8 and 8.6%.

6.1 Introduction

To help inform public policy and to abate threats posed by climate change,

the United Nations Framework Convention on Climate Change (UNFCCC) requires

all countries to regularly report the state of their forests and assessments of C stocks

(2007). The Reducing Emission from Deforestation and forest Degradation (REDD-

plus) monetary scheme provides incentives for enhancing and conserving carbon (C)

sequestration. However, definitions of additive C sequestration differ between

countries and many challenges remain to estimating, monitoring and verifying C

stocks in forest ecosystems.

Direct measurement of tree biomass is seldom feasible. The most widely

used indirect method involves use of allometric equations, that are usually based on

biometrics such as diameter at breast height (DBH), tree height (Somogyi et al.,

2007) and specific gravity (SG) of wood (Chave et al., 2005; Chave et al., 2014).

Allometric equations are best developed and tested for individual species, forest

types and sites (Petersson et al., 2012). Comprehensive databases for species-specific

biomass and volume equations are available for North America (Ter-Mikaelian and

Korzukhin, 1997) and Europe (Levy et al., 2004; Muukkonen, 2007). General

allometric equations are available for pan-tropical trees (Chave et al., 2014), by

taxonomic grouping (Chojnacky et al., 2014), and for individual countries (Paul et

al., 2016). For the Himalayan region, a few allometric equations have been

developed for species in the western (Garkoti, 2008) and central (Negi et al., 1983;

Rana et al., 1989) areas based on limited sample size and some for small diameter

trees (Singh et al., 2011).


Page | 143

Customarily, timber has been the main forest product and traditional forest

inventories were designed to document timber volume, growth rates and mortality.

Consequently, there is considerable data on timber volume, and many countries have

resorted to converting volume data to biomass by using a biomass expansion factor

(BEF). Even here many regions and countries lack BEF for their forest types with

resultant increases in uncertainties of biomass estimations (Teobaldelli et al., 2009).

At the same time, biomass estimation based on tree volumes can result in serious

overestimation, e.g. by 19% (Moundounga Mavouroulou et al., 2014). Given the lack

of a substantive basis for estimating biomass in the eastern Himalayas, we harvested

144 trees with DBH ranging from 10 to 77 cm from different forest types along an

altitudinal transect, with the aim to develop allometric equations. We also conducted

a forest inventory with the aim to apply our equations to estimate biomass and C

stocks for the different forest types.

6.2 Material and methods

6.2.1 Study site

Our sites are located in the eastern foothills of the Himalayas in Bhutan.

Initial scoping was completed during the winter of 2012 and 2013 along a transect

starting from Phuentsoling at an elevation of 317 m (N 26° 51´, 89° 23´ E) to Gedu

top, at an elevation of 3300 m (N 26° 59´, 89° 32´ E). The foothills experience a

tropical climate with mean annual rainfall as high as 4600 mm and mean annual

temperature of 22.9 °C (2009, Department of Hydro-Met Services, Bhutan). In the

mid-altitudes between 2000 to 3000 m a.s.l., the annual precipitation is about 3500

mm, with a maximum summer temperature of 29 °C and minimum winter

temperature of 3 °C in December (Wangda et al., 2009).

6.2.2 Forest inventory and zonation

Along the study transect, 20 inventory plots (30 m × 30 m) spaced out at

150 m altitudinal intervals were surveyed. Every individual tree (>10 cm diameter at

breast height, DBH) was identified and measured for diameter at 1.3 m above

ground, from the uphill side. Tree height was estimated with a Haglöf Vertex


Page | 144

hypsometer and transponder. Within each 30 m × 30 m plot, three (5 m × 5 m)

understorey vegetation plots were used to measure vegetation with diameter < 10 cm

DBH. Additionally, 3 (1 m × 1 m) plots were randomly established within every 30

m × 30 m to collect and the surface plant residue measured. From the initial forest

inventory, the vegetation was zoned into tropical forest (TF, 317 − 900 m a.s.l), sub-

tropical forest (STF, 900 − 1870 m a.s.l), warm temperate broadleaf forest (WTBLF,

1870 − 2450 m a.s.l), cool temperate broadleaf forest (CTBLF, 2450 − 3000 m a.s.l),

and cold temperate forest (CTF, 3000 − 3300 m a.s.l., Table 6.1,(Ohsawa, 1987).

Table 6.1 Characteristics of forest along the altitudinal gradient.

Forest

Type

Altitude

(m a.s.l.)

No.

Species

Stem

density

(ha-1)

Max tree

Height (m)

Max tree

DBH (cm)

BA

(m2 ha-1)

TF 317−900 33 313 34 74 17.8

STF 900−1870 54 383 50 101 30.1

WTBLF 1870−2450 42 433 43 106 41.1

CTBLF 2450−3000 47 824 43 101 46.3

CT 3000−3300 10 511 38 120 66.6

TF, tropical forest; STF, subtropical forest; WTBLF, warm-temperate broadleaf forest;

CTBLF, cool-temperate broadleaf forest; CTF, cold temperate forest; No. of species, number

of tree species surveyed for each forest type; stem density, number of trees per ha; BA, basal

area of trees per ha (m2 ha-1).

A total of 137 tree species were identified from the study site during the

forest inventory. Not all tree species could be harvested and measured to develop the

biomass equation due to time and monetary constraints. For each forest type, trees

were grouped into diameter classes of 10 cm intervals, starting from 10 cm up to the

biggest tree, and five trees from each diameter class from the different forest types

were randomly selected using the random function in Microsoft excel 2007 (144

trees harvested in total).


Page | 145

6.2.3 Tree biomass data collection

In the field, trees to be harvested were identified, the GPS locations and

altitude recorded and DBH and height measured. Thereafter each tree was felled and

a randomized branch sampling (RBS) methodology adopted to estimate AGB (Good

et al., 2001; Gregoire et al., 1995; Somogyi et al., 2007; Valentine et al., 1984). This

methodology requires only selected portions of the tree to be measured and weighed,

and is more suitable to remote areas and large sample sizes. RBS was first adopted

by (Jessen, 1955) for selecting sub-samples based on probability proportional to size,

and is able to estimate many characteristics of trees, such as aboveground woody dry

matter (Valentine and Hilton, 1977). A Haglöf 24"Complete Increment borer (5.15

mm diameter) was used to extract a wood core from 1.3 m and at one third (centroid)

height of the tree. From each section above and below the centroid, two 4 cm thick

discs were harvested from heights determined by random numbers generated and

height of the tree. The disc thickness was measured with a digital Vernier calliper at

four evenly spaced points on the circumference of the disc and recorded. From the

RBS selected pathways, larger diameter branch segments were further sampled to

reduce the amount of samples for transportation and oven drying. Each selected path

or branch segment was measured (distance between two nodes). A c. four cm thick

disc was extracted at the midpoint of the branch segment and the exact thickness

measured with a Vernier calliper. With subsequent sampling, as selected branches

became sufficiently small, they were clipped, weighed and bagged. Small epicormic

or dwarf branches that occurred at nodes or between two successive nodes were

clipped, weighed and a subsample bagged. All wood discs as well as terminal shoots

were oven dried at 105 °C to constant weights which were used to calculate mass of

the tree:

Mass of segment (g) = length of segment (cm)×mass of sample (g)

width of sample disc (cm) Equation 6. 1

To measure the C content in the tree biomass samples, dried wood cores

and leaves were ground into a fine powder with a mortar and pestle. About 3 mg of

samples were weighed into tin cups, folded and combusted at 1000 °C in a Thermo

Delta V isotope ratio mass spectrometer coupled to ConfloIV and FlashHT


Page | 146

peripherals (Thermo Fisher, Bremen, Germany). Precision of the % C was below

0.17%.

6.2.4 Measurement of specific gravity of wood cores

Wood cores extracted at 1.3 m above ground for each harvested tree were

used to measure the wood specific gravity (SG) of individual trees. Air dried cores

were initially soaked in a beaker of deionized water to ensure adequate swelling.

Thereafter, a beaker of deionized water was placed on a top pan GX-2000 digital

balance with a precision of 0.01 g. The balance was zeroed, and wood core

submerged using a needle. The weight of water displaced is equal to volume of the

submerged wood core (1 g of water is equal to 1 cc at 4 °C, corrected to 18 °C room

temperature, where density of water is 0.998 g cc-1). Thereafter, the wood corers

were oven dried at 105 °C to constant weight and specific gravity of wood calculated

as:

Specific gravity of wood (SG) = oven dried weightvolume of displaced water

Equation 6.2

6.2.5 Calculation of biomass in the understorey vegetation

To estimate the forest biomass of trees with DBH less than 10 cm, but

greater than 1.3 m height, three 10 m2 plots were randomly chosen in each of the

forest zones. All the vegetation was destructively harvested, weighed and subsamples

collected. For trees smaller than 1.3 m in height, three 1 m2 plots were established

within the inventory plots and all biomass harvested, weighed and sub samples

bagged and subsequently oven dried at 105 °C to constant weight. Oven-dry mass of

the different vegetation class subsamples were used to calculate total dry biomass:

Understory biomass = total biomass (FW)× subsample biomass (ODW)subsample biomass (FW)

Equation 6.3


Page | 147

6.2.6 Models and statistical analysis

Scatterplots for AGB data and tree parameters such as DBH, height and

wood SG were used to detect trends in relationships. AGB data and tree parameters

were log transformed (Supplement 6.1) so that standard least-square regression could

be used to develop best-fit models. Transformation helped linearize relationships and

equalize variance over the range of measured biomass (Basuki et al., 2009; Picard et

al., 2012). The following models were tested to determine the best allometric

equation for different forests along the altitudinal gradient.

Model 1: log AGB = α + β1 log (DBH2) Equation 6.4

Model 2: log AGB = α + β1 log (DBH2) + β2 log (SG) Equation 6.5

Model 3: log AGB = α + β1 log (DBH2) + β2 log (SG) + β3 log (H) Equation 6.6

Model 4: log AGB = α + β1 log (DBH2 × H) Equation 6.7

Model 5: log AGB = α + β1 log (DBH2 × SG) Equation 6.8

Model 6: log AGB = α + β1 log (DBH2 × SG × H) Equation 6.9

where log is the logarithm to the base 10, AGB the dry weight of total tree

aboveground biomass, DBH is diameter of tree at breast height, SG is the specific

gravity of wood for the respective tree species and α, β1, β2 and β3 are parameters to

be estimated.

IBM SPSS (Armonk, NY, USA) was used to conduct regression analyses,

with log transformed AGB data as the response variable and log DBH2, log (DBH2 ×

height), log (DBH2 × SG) and log (DBH2 × height × SG) used as predictor variables.

To assess the fit of each biomass equations for the separate forest types the a) p-

value, b) adjusted r2 values c) Standard error of the estimate (SEE), d) Akaike

Information Criterion (AIC) based on likelihood and number of model parameters

were calculated (Akaike, 1974; Mazerolle, 2004):


Page | 148

AIC = − 2*ln (likelihood) + 2K

AIC = n*ln �RSSn� + 2K Equation 6.10

where K is the number of parameters + 1, ln = natural logarithm, n = number of

observations, and RSS = residual sum of squares. AIC requires bias adjustment for

small sample sizes (rule of thumb: if ratio of n/K < 40):

AICc= −2* ln(likelihood) + 2K + �2K(K + 1)n −K−1

� Equation 6.11

To interpret the data, Akaike weights which are the normalized relative

likelihood were calculated with the following formula.

wi =exp (− 0.5 × ∆i)

∑ exp (− 0.5 × ∆r)R

r=1

Equation 6.12

∆i= (AICi − AICmin) and AICi is AIC for model i, exp (-0.5 × ∆r) is the sum of relative

likelihoods of all candidate models.and e) average deviation which was calculated as

the absolute difference between estimated dry biomass and measured dry biomass

expressed as the proportion of the measured dry biomass (Basuki et al., 2009; Chave

et al., 2005).

S (%) = �∑ �yi − 𝑦𝑖�𝑦𝑖

ni=1 �× 100

n Equation 6.13

where S is the average deviation, yi is the estimated dried weight, yi is the measured

dry weight, n is the number of observations.

While log transformation of data satisfies statistical requirements and

simplifies model building, it can introduce systematic bias into the calculation

(Sprugel, 1983). That bias can be counteracted with a correction factor while back

transforming the equation for implementation:


Page | 149

CF = exp �SEE2

2� Equation 6.14

where CF is correction factor, SEE is based on natural logarithms (to convert a

natural antilog to a base = 10 antilog, multiply by the conversion factor of 2.303

before taking the antilog). Errors involved are usually less than 10% (Sprugel, 1983).

Given the limited data sets for each forest type, 3-fold cross validation was

used to evaluate the predictive performance of each of the models. The root mean

square error of the training and validation data was calculated and averaged for each

model. The model mean relative prediction error was determined by calculating the

percent bias (% bias) for the validation data set using the equation (Arevalo et al.,

2007).

% Bias = �∑ Wi − Ŵi Wi

ni=1 �× 100

n Equation 6.15

where Wi is the observed AGBs, Ŵi the estimated AGB and n is the number of

observations. After the models were validated, the entire data set was used build the

model as we had a limited data set.

6.2.7 Comparison of model selected for each forest type with previously

published equations

We chose AGB equations from those proposed for taxonomic groupings by

Chojnacky et al. (2014). We selected those most similar to the taxonomic grouping

for the present study to make comparisons. We used their α and β1 coefficients for

the model ln (biomass) = α + β1ln (DBH) and chose Abies forest for CTF (α =

−3.1774, β1 = 2.6426), Aceraceae hardwood forest for CTBLF (α = −2.0470, β1 =

2.3852), Magnolia forest for WTBLF (α = −2.5497, β1 = 2.5011), Cornaceae,

Ericaceae, Lauraceae, Plantanaceae, Rosaceae and Ulmaceae forest for STF (α =

−2.2118, β1 = 2.4133), and Fagaceae evergreen forest for TF (α = −2.2198, β1 =

2.4410). Additionally we considered the global pan tropical moist forest model from

(Chave et al., 2005), AGB = 0.0509 × ρDBH2H, where ρ = specific gravity of wood,

DBH = diameter at breast height of tree in cm and H = height of tree in m.


Page | 150

6.3 Results

6.3.1 Carbon concentrations in overstorey tree wood, leaves and specific gravity

of wood

In order to correctly estimate C stock for the five forest types and

parameterize the models, we measured C concentrations in wood and leaves as well

as the specific gravity of wood along the altitudinal gradient. Wood C concentration

for individual tree trunks ranged from 39.59 to 47.85% (Supplement 6.2), but when

the average was taken, C concentration for all forest types were similar (46.02 to

46.81%, Table 6.2). Individual leaf C concentrations varied from 42.99 to 56.92%

(Supplement 6.2), and mean leaf C concentrations (48.70 to 52.17%, Table 6.2) were

significantly greater for the CTF than for other forest types. Average specific gravity

of wood varied between 0.48 and 0.63% for the different forest types (Table 6.2) and

was least for CTF at the highest altitude.

Table 6.2 Carbon content (Mean ± SD) on mass (%) basis and specific gravity of

wood in the overstorey biomass from different forest zones in Bhutan Forest zone Wood C (%) Leafs C (%) Leafsh C (%) Wood SG (g cm-3)

TF 46.02a ± 1.88 48.99a ± 2.27 48.70a ± 2.44 0.57a ± 0.16

STF 46.81a ± 1.36 50.23a ± 1.40 49.51a ± 1.66 0.63a ± 0.11

WTBLF 46.08a ± 1.52 49.90a ± 2.53 49.23a ± 2.67 0.53b ± 0.12

CTBLF 46.32a ± 0.99 49.92a ± 2.99 49.51a ± 2.22 0.56ab ± 0.09

CTF 46.78a ± 1.13 52.02b ± 0.63 52.17b ± 1.21 0.48b ± 0.07

s = sun leaves, sh = shade leaves. Different letters within each column indicate significant

difference between the forest types (P < 0.05).


Page | 151

6.3.2 Model selection

We estimated AGB using all models (equations 6.4 to 6.9) for the different

forest types. We chose the most appropriate model based on AIC, RMSE, r2, p-

values, absolute average deviation and coefficients of models for each of the forest

types (Table 6.3).

6.3.3 Model selection for the tropical forest

AGB predictions from Model 3 for TF had the lowest average deviation

from the observed AGB (Table 6.3). However, Model 4 (Fig. 6.1a) had the lowest

AIC, RMSE, highest r2 (0.918) and second lowest average deviation amongst the

models (Table 6.3). Cross validation suggests that Model 4 consistently performs

better compared to other models, with least RMSE and greatest r2 for the training and

test data sets and lowest estimation % bias (−14.6%, Table 6.4).

6.3.4 Model selection for the sub-tropical forest

Model 6 (Fig. 6.1b) had the least AIC and greatest r2 (0.907) values for STF.

In comparison, Model 4 had slightly greater AIC and RMSE values but lesser

average deviation to Model 6 (Table 6.3). Furthermore, when the models were cross

validated, Model 4 consistently performed better with lower RMSE values as well as

more stable RMSE and r2 value for the training and test data sets (Table 6.4). Wood

SG, a variable considered in Model 6, may not always be collected during forest

inventory; therefore Model 4 is recommended for estimation of AGB for STF.

Akaike weights of Models 6 and 4 were calculated to be 0.483 and 0.299,

respectively. This suggests Model 6 is (0.483 ÷ 0.299) 1.6 times more likely to make

better estimation than Model 4 for AGB. Nonetheless, Akaike weights for Model 4

are well within the confidence set of Model 6 (values > than 10% of the weight of the

best model) and thus Model 4 is a valid alternative.


Page | 152

6.3.5 Model selection for the warm tropical broadleaved forest

Model 3 had the least RMSE, AIC values and the greatest r2 values for the

WTBLF (Table 6.3). However, Model 1 had comparable RMSE, AIC and r2 values

to Model 3. Akaike weights for Models 3 and 1 were 0.541 and 0.088, respectively.

Model 3 is thus 6.1 times more likely to provide reasonable estimations than Model

1. Model 1 is still within the confidence sets of candidates models (> 10% of the

highest Akaike weights) and therefore valid as an alternative model (Symonds and

Moussalli, 2011) especially if SG of wood is not available.

6.3.6 Model selection for the cool temperate broadleaved forest

Based on the lowest AIC, RMSE, highest r2 values and relatively low

average deviation, Model 5 was the best predictor model for CTBLF; however, SG

of wood is required for this model. Model 1 was the best model without the SG

parameter, and has comparable values of AIC, RMSE, r2 and average deviation to

Model 5. Model 5 (with DBH and wood specific gravity) is 4.4 time more likely to

be a better explanation for forest biomass than tree DBH alone (Akaike weight of

Model 5 to Model 1: 0.390 ÷ 0.089). However, the Akaike weight for Model 1 is

within the confidence set of the best model and therefore valid as an alternative when

wood SG is not available.

6.3.7 Model selection for the cold temperate forest

Measured AGB data did not deviate much from a simple quadratic model

prediction; hence Model 1 which uses only log (DBH2) was the best AGB predictor

for CTF with the lowest AIC, RMSE, relatively low average deviation of 17.8% and

highest r2 values compared to other models (Table 6.3). Cross validation revealed

that Model 1 outperforms the other models (lowest RMSE and highest r2 values

Table 6.4). Model 2 (including SG of wood) had comparable AIC, RMSE and r2

values to Model 1 and slightly lower average deviation than Model 1. Inclusion of

SG in Model 2 resulted in similar predictions to the simple quadratic Model 1 (Fig.

6.1e), which is likely due to little diversity of tree species and similar SG of wood for

the trees in CTF.


Page | 153

6.3.8 Model selection for the entire forest

When all sampled trees were used to construct combined AGB equations,

Model 3 was the best predictor model, based on lowest AIC, RMSE and average

deviation (Table 6.3). Cross validation results also showed lowest percent bias for

Model 3 (Table 6.4). However Model 3 required more variables like DBH, H and SG

of wood to be entered as separate effect variables to estimate AGB. Model 4 with

(DBH2*H) entered as a single effect variables to estimate the AGB had comparable

model selection criteria values and is recommended when SG of individual trees is

not available.

Comparisons of selection parameters of best-predictor-models for the entire

forest (Models 3 and 4) with models selected for the specific forest types were

similar for most forest types, except for the CTF at the highest altitude, where the

whole-forest Model 4 compared to the CTF-specific Model 1. More specifically,

Model 3 (whole-forest selection) was also the best fit model for TF and WTBLF

(Fig. 6.1a&c) and for CTBLF, Model 3 (whole-forest selection) had comparable

model selection criteria values to Model 1 (CTBLF selection). In the STF, Model 3

(whole-forest selection) and Model 6 (STF selection) had identical RMSE and

similar r2, CF and absolute average deviation values, but a higher AIC value (Table

6.3).


Page | 154

Table 6.3 List of models developed for estimation of aboveground tree biomass for

the different forest types in Bhutan.

Model Tropical forest (26) α β1 β2 β3 RMSE r2 AICc CF S (%)

1 log AGB = α + β1 log (DBH2) -0.922** 1.182**

0.171 0.902 -89.4 1.08 32.7

2 log AGB = α + β1 log (DBH2) + β2 log (SG) -0.848** 1.181** 0.266ns 0.170 0.904 -88.4 1.079 31.8

3 log AGB = α + β1 log (DBH2) + β2 log (SG) + β3 log (H) -1.211** 0.943** 0.29ns 0.876* 0.157 0.917 -90.8 1.067 27.4

4 log AGB = α + β1 log (DBH2*H) -1.308** 0.928** 0.157 0.918 -93.9 1.067 29.0

5 log AGB = α + β1 log (DBH2*SG) -0.334ns 1.083** 0.212 0.850 -78.3 1.126 38.5

6 log AGB = α + β1 log (DBH2*SG* H) -0.889** 0.885** 0.177 0.895 -87.6 1.086 28.8

Model Sub- tropical forest (28) 1 log AGB = α + β1 log (DBH2) -0.434ns 1.075** 0.160 0.889 -100.4 1.069 30.1

2 log AGB = α + β1 log (DBH2) + β2 log (SG) -0.44ns 1.111** 0.538ns 0.158 0.896 -100.8 1.068 27.9

3 log AGB = α + β1 log (DBH2) + β2 log (SG) + β3 log (H) -0.514* 0.895** 0.484ns 0.587ns 0.149 0.906 -102.3 1.062 25.5

4 log AGB = α + β1 log (DBH2*H) -0.516* 0.784** 0.148 0.904 -104.5 1.060 27.1

5 log AGB = α + β1 log (DBH2*SG) -0.351ns 1.122** 0.164 0.887 -100.1 1.073 30.3

6 log AGB = α + β1 log (DBH2*SG* H) -0.473* 0.812** 0.149 0.907 -105.4 1.060 27.8

Model Warm temperate broadleaved forest (31) 1 log AGB = α + β1 log (DBH2) -1.02** 1.224** 0.109 0.959 -135.2 1.031 19.3

2 log AGB = α + β1 log (DBH2) + β2 log (SG) -0.878** 1.210** 0.347ns 0.105 0.962 -136.0 1.029 18.0

3 log AGB = α + β1 log (DBH2) + β2 log (SG) + β3 log (H) -0.968** 1.039** 0.441* 0.528* 0.098 0.967 -138.8 1.025 15.9


5 log AGB = α + β1 log (DBH2*SG) -0.433** 1.141** 0.131 0.941 -123.8 1.046 22.9


Model Cool temperate broadleaved forest (30)

1 log AGB = α + β1 log (DBH2) -0.887** 1.189** 0.135 0.923 -117.9 1.049 24.8

2 log AGB = α + β1 log (DBH2) + β2 log (SG) -0.587* 1.153** 0.738ns 0.128 0.931 -119.7 1.044 23.7

3 log AGB = α + β1 log (DBH2) + β2 log (SG) + β3 log (H) -0.739ns 1.103** 0.744ns 0.252ns 0.129 0.929 -117.4 1.045 23.6


5 log AGB = α + β1 log (DBH2*SG) -0.397* 1.125** 0.128 0.930 -120.9 1.049 24.1


Model Cold temperate forest (25) 1 log AGB = α + β1 log (DBH2) -1.113** 1.284** 0.102 0.956 -111.9 1.027 17.8

2 log AGB = α + β1 log (DBH2) + β2 log (SG) -1.122** 1.279** -0.078ns

0.104 0.954 -109.5 1.028 17.6

3 log AGB = α + β1 log (DBH2) + β2 log (SG) + β3 log (H) -1.191** 1.245** -0.097ns 0.142ns 0.106 0.952 -106.6 1.030 17.7

4 log AGB = α + β1 log (DBH2*H) -1.558** 1.035**

0.112 0.947 -107.2 1.033 20.2



Model Entire Forest (140)


Page | 155

1 log AGB = α + β1 log (DBH2) -0.901** 1.198** 0.147 0.917 -525.6 1.059 42.0

2 log AGB = α + β1 log (DBH2) + β2 log (SG) -0.807** 1.198** 0.355**

0.142 0.923 -533.8 1.055 27.0

3 log AGB = α + β1 log (DBH2) + β2 log (SG) + β3 log (H) -0.900** 1.128** 0.346** 0.249 0.141 0.924 -534.8 1.055 26.7




DBH in cm and H in m; ** p < 0.01; *p < 0.05 and ns = non-significant, number in

parenthesis after each forest type is the total number of tree harvested, the best performing

and recommended models (e.g. if the SG parameter is unavailable) for each forest type are

depicted in bold. RMSE = root mean square error, r2 = coefficient of determination, AIC =

Akaike information criterion, CF = correction factor and S (%) = absolute average deviation

Table 6.4 Mean RMSE, r2 and % bias for training and validation data set for the

different models used for various forest types.

Training Test Model no. TF Biomass equation n RMSE r2

n RMSE r2 Bias (%)

1 log AGB = α + β1 log (DBH2) 18 0.171 0.89 8 0.173 0.87 -17.2

2 log AGB = α + β1 log (DBH2) + β2 log (SG) 18 0.172 0.89 8 0.181 0.86 -17.0

3 log AGB = α + β1 log (DBH2) + β2 log (SG) + β3 log (H) 18 0.156 0.91 8 0.167 0.86 -

15.1

4 log AGB = α + β1 log (DBH2*H) 18 0.155 0.91 8 0.150 0.89 -14.6

5 log AGB = α + β1 log (DBH2*SG) 18 0.213 0.83 8 0.214 0.78 -28.2

6 log AGB = α + β1 log (DBH2*SG* H) 18 0.177 0.88 8 0.176 0.82 -18.2

STF Biomass equation 1 log AGB = α + β1 log (DBH2) 18 0.159 0.89 9 0.154 0.87 -

15.6

2 log AGB = α + β1 log (DBH2) + β2 log (SG) 18 0.159 0.89

9 0.161 0.86 -11.0

3 log AGB = α + β1 log (DBH2) + β2 log (SG) + β3 log (H) 18 0.153 0.90 9 0.163 0.86 -

10.0

4 log AGB = α + β1 log (DBH2*H) 18 0.149 0.90 9 0.153 0.87 -12.9

5 log AGB = α + β1 log (DBH2*SG) 18 0.163 0.88 9 0.164 0.85 -13.0


WTBLF Biomass equation 1 log AGB = α + β1 log (DBH2) 21 0.109 0.96 10 0.104 0.96 -6.4 2 log AGB = α + β1 log (DBH2) + β2 log (SG) 21 0.106 0.96 10 0.104 0.96 -5.8

3 log AGB = α + β1 log (DBH2) + β2 log (SG) + β3 log (H) 21 0.099 0.97 10 0.101 0.96 -4.7

4 log AGB = α + β1 log (DBH2*H) 21 0.111 0.96 10 0.109 0.96 -6.7 5 log AGB = α + β1 log (DBH2*SG) 21 0.131 0.94 10 0.129 0.94 -9.8 6 log AGB = α + β1 log (DBH2*SG* H) 21 0.106 0.96 10 0.105 0.96 -6.2

CTBLF Biomass equation 1 log AGB = α + β1 log (DBH2) 20 0.130 0.93 10 0.123 0.93 -


Page | 156

10.0

2 log AGB = α + β1 log (DBH2) + β2 log (SG) 20 0.126 0.93 10 0.124 0.93 -8.8

3 log AGB = α + β1 log (DBH2) + β2 log (SG) + β3 log (H) 20 0.126 0.93

10 0.120 0.93 -8.9

4 log AGB = α + β1 log (DBH2*H) 20 0.136 0.92

10 0.129 0.92 -11.1

5 log AGB = α + β1 log (DBH2*SG) 20 0.126 0.93 10 0.123 0.93 -9.6 6 log AGB = α + β1 log (DBH2*SG* H) 20 0.128 0.93 10 0.124 0.93 -9.7

CTF Biomass equation 1 log AGB = α + β1 log (DBH2) 17 0.102 0.95 8 0.103 0.94 -5.5 2 log AGB = α + β1 log (DBH2) + β2 log (SG) 17 0.100 0.95 8 0.091 0.95 -5.6


4 log AGB = α + β1 log (DBH2*H) 17 0.110 0.95 8 0.107 0.94 -6.7 5 log AGB = α + β1 log (DBH2*SG) 17 0.127 0.93 8 0.126 0.92 -9.1


Entire forest Biomass equation 1 log AGB = α + β1 log (DBH2) 91 0.147 0.92 45 0.146 0.92 -5.7 2 log AGB = α + β1 log (DBH2) + β2 log (SG) 90 0.143 0.92 43 0.143 0.92 -5.2


4 log AGB = α + β1 log (DBH2*H) 91 0.154 0.91 45 0.154 0.91 -6.6 5 log AGB = α + β1 log (DBH2*SG) 91 0.165 0.90 44 0.166 0.89 -7.5 6 log AGB = α + β1 log (DBH2*SG* H) 91 0.163 0.90 44 0.164 0.90 -6.9

DBH in cm and H in m; RMSE = root mean square error, r2 = coefficient of

determination, n = sample size, Bias (%) = percentage bias.


Page | 157

Fig. 6.1 The relationship between observed and predicted total above ground tree

biomass with tree DBH for a) tropical forest, b) Sub tropical forest, c) Warm

temperate broadleaf forest, d) Cool temperate broadleaf forest and e) Cold temperate

forest. Filled diamond = measured total tree biomass, filled square = best

predictor total tree biomass model and filled triangle = proposed alternative total

tree biomass model.


Page | 158

6.3.9 Model comparison to published equations

When the equations from Chojnacky et al. (2014) and Chave et al. (2005)

were applied to our data, the predicted values consistently overestimated AGB for TF

at the lowest altitude and underestimated AGB for CTF at the highest altitude.

Confidence interval at 95%, upper and lower values of the mean were wider for AGB

prediction from Chojnacky et al. (2014) and Chave et al. (2005) for the TF and

inconsistent for the other forest types (Table 6.5). Average deviation for predictions

of AGB with selected models from different forest was lower than prediction by

other published AGB equations, except for prediction by Chojnacky et al. (2014) for

WTBLF and Chave et al. (2005) for CTBLF. When comparisons between AGB

estimates with Model 4 for entire forest, Chojnacky et al. (2014) and Chave et al.

(2005) were made, the AGB averages, CI as well as average deviation are closer to

the observed AGB.

Table 6.5 Confidence interval (CI) of the measured mean AGB and average

deviation of estimated AGB using models of best fit developed for the different

forest types in the current study and compared to models by Chojnacky et al. (2014)

and Chave et al. (2005). n Observed Predicted Chave Chojnacky

TF (Model 4)

Mean AGB (kg) 27 899.6 878.2 1169.6 1048.2

95 % lower limit of mean AGB (kg) 538.2 504.5 621.2 643.3

95 % upper limit of mean AGB (kg) 1260.9 1251.9 1717.9 1453.2

Average deviation (%) - 29.0 39.8 41.3

STF (Model 4)

Mean AGB (kg) 29 1516.2 1445.6 1693.4 1250.5




WTBLF (Model 1)

Mean AGB (kg) 31 1061.0 1033.2 1107.3 1042.4




CTBLF (Model 1)

Mean AGB (kg) 30 1091.1 1035.9 1072.4 1060.8





Page | 159

CTF (Model 1)

Mean AGB (kg) 25 1342.6 1314.4 779.2 953.6




Entire forest (Model 4)

Mean AGB (kg) 140 1178.8 1129.8 1170.0 1073.2




6.3.10 Aboveground overstorey and understorey biomass of the different forest

types

Total AGB (estimated by random branch sampling) and C content of trees >

10 cm DBH increased with increasing altitude of the forest. Tropical forest at the

lowest altitude had AGB of 108 Mg ha-1and a C stock of 49.8 Mg ha-1. Aboveground

tree biomass increased to 407.2 Mg ha-1 while C stock increased to 190.5 Mg ha-1 for

CTF at the highest altitude (Table 6.6). Similarly, for trees less than 10 cm, the AGB

and C stock increased with altitude up to the mid-altitudes and then became more

variable. In contrast, the aboveground biomass and C stock of understorey vegetation

(< 1.3 m height) decreased with increasing altitude of the forest. The amount of

deadwood across the forest types is similar, while the quantity of leaf litter and C

stocks increased up to the second highest altitude (CTBLF) and dropped slightly for

the highest altitude (CTF). Nonetheless, the total AGB as well as the C stocks

increased with increasing altitude of the forest.


Page | 160

Table 6.6 Total measured aboveground biomass and carbon from various biomass

components for different forest types.

Forest

Trees Trees Shrubs

Deadwood Litter Total (> 10 cm DBH)

(< 10 cm DBH)

(< 1.3 m height)

Total biomass (Mg ha-1)

TF 108.24 ± 77.0 6.65a ± 2.2 7.64a ± 6.6 2.64a ± 1.7 1.48a ± 0.5 126.65

STF 175.67 ± 52.4 9.94ac ± 2.7 6.12a ± 2.2 2.49a± 1.7 2.35a ± 1.2 196.57

WTBLF 197.64 ± 45.9 34.77bd ± 14.2 6.55a ± 1.8 3.78a ± 3.0 4.18b ± 2.0 246.92

CTBLF 313.77 ± 98.1 15.69ad ± 1.7 2.41b ± 1.6 2.49a ± 1.6 5.47b ± 2.2 339.83

CTF 407.23 28.62cd ± 7.7 2.97ab ± 1.3 2.24a ± 0.7 3.48ab ± 0.6 444.54 Total biomass carbon (Mg C ha-1)

TF 49.81 ± 35.4 2.73a ± 0.9 3.21a ± 2.8 1.09a ± 0.6 0.62a ± 0.2 57.46

STF 82.23 ± 24.5 4.2ac ± 1.1 2.64a ± 1.1 1.11a ± 0.7 1.06a ± 0.5 91.24

WTBLF 91.07 ± 21.1 15.1b ± 6.1 2.85a ± 0.8 1.75a ± 1.4 1.90b ± 0.9 112.67

CTBLF 145.34 ± 45.4 7.14ab ± 0.8 1.11b ± 0.7 1.19a ± 0.8 2.59b ± 1.0 157.37 CTF 190.5 13.05bc ± 3.5 1.35ab ± 0.6 1.09a ± 0.3 1.66ab ± 0.2 207.65

Different letters within each column indicate significant difference between the forest types

(P < 0.05).

6.4 Discussion

6.4.1 Biomass carbon concentration and aboveground biomass in the different

forest types

Accurate measurements of C concentrations in different components of the

tree biomass are important to limit over- or underestimation of forest C stocks

(Thomas and Martin, 2012). Although C concentrations in tree wood for different

forest types were not significantly different, C concentration slightly increased with

altitude, as did C concentrations for overstorey leaves and understorey vegetation

(Tashi et al., 2016). Wood C concentrations have been shown to vary with biomes as

well as with tree species, e.g. between 41.9 to 51.6% for tropical forest, between 45.7

to 60.7% for subtropical forest and between 43.4 to 55.6% for temperate forest

(Thomas and Martin, 2012). Average wood C concentrations for all forest types in

the present study were c. 4% lower than the 50% value used for most regional and

global assessment of C stocks (Houghton et al., 2000; Saatchi et al., 2011). We were


Page | 161

able to reduce the over-estimation of C stock by c. 6.8 to 8.6% by using species- and

tissue-specific C concentration, rather than the assumed 50% C in biomass

(Supplement 6.3). In contrast to C concentration, AGB estimates for the different

forests increased significantly with altitude (108 to 407 Mg ha-1), and is comparable

to values reported from other studies, including for lodgepole pine biomass stocks in

Canada (172 Mg ha-1 to 425 Mg ha-1, (Monserud et al., 2006), for different land form

types on Barro Colorado Island (Chave et al., 2003) and for all major forest types in

China (Fang et al., 1998). The increased biomass and C stocks with increasing

altitude of the forests documented here was mainly due to the increase in BA and

density of the forest with increasing altitude.

6.4.2 Aboveground tree biomass model selection

While a good range of tree diameters in each forest type along the altitudinal

gradient were considered (10 to 77 cm), all models remain based on limited data and

we used 3-fold cross validation to address this limitation. The goodness-of-fit

statistics for the training and test data are similar across all forest types for the

different AGB models considered (Table 6.4) and therefore valid for AGB model

building. The best AGB predictor models for all forests included tree DBH, height

and wood SG. Tree height is an indicator of site quality (Nogueira Júnior et al., 2014;

Skovsgaard and Vanclay, 2008) and tree diameter together with height (DBH2 ×

height) defines the structural patterns of most trees (Picard et al., 2012). Inclusion of

wood SG led to improvement in the AGB estimation, generally captured the

variability of the measured data better, and has been shown to be especially

important for contrasting sites containing many different forest tree species (Basuki

et al., 2009; Chave et al., 2005; Chave et al., 2014). Wood SG did not correlate with

either tree size (DBH and height) or C concentrations in wood as reported by Martin

and Thomas (2011). In contrast, significant correlations between forest type and SG

of wood suggest that SG is more determined by forest composition as reported by

Baker et al. (2004). However, wood SG may not always be available or collected

during a conventional forest inventory. As more data on wood SG is increasingly

available from compilations such as Zanne et al. (2009) and other individual studies,

allometric equations with wood SG will become more relevant.


Page | 162

When models included only DBH2 as the effect variable like the

recommended model for WTBLF, CTBLF and CTF, AGB may be potentially

overestimated, especially for larger trees (Chave et al., 2005). In the present study,

we did not overestimate AGB, as we restricted our biomass models to the range of

DBH used to construct the equations (Banaticla et al., 2005). On the contrary, the

chosen models used to predict AGB for the different forest types underestimated the

AGB between 0.78 to 4.75%. Inclusion of tree height in the model without

considering crown diameters could have led to the slight underestimation of the

biomass of trees (Goodman et al., 2014).

Notably in the three higher altitude forest types, the simplest model with

only DBH2 as the effect variable showed the best fit to the data. Still, the form of the

models chosen for the whole forest performed similarly well for the separate forest

types based on all selection parameters and had similar bias, with the exception for

the highest altitude forest (CTF). Here, the two models containing the SG parameter

(Model 2 as best fit for this forest and model 3 as best fit for the whole forest) were

very close in their selection parameters; however model 4 (whole forest) performed

worse than model 1 (best fit for this forest). The best fit of the simplest model may

have been influenced by the lower variability of measured data in the three higher

altitude forests, and the small data set, especially for the highest altitude forest

(CTF).

6.4.3 Comparison of models to various other models

Various studies have proposed that for accurate estimation of AGB, site-

specific species equations must be developed (Basuki et al., 2009; Cairns et al.,

2003). Our site-specific models provided better estimates of AGB for lower and high

altitude forests compared to general AGB models proposed by Chave et al. (2005)

and Chojnacky et al. (2014). For mid-altitude forests, the deviations between

observed AGB and estimated AGB between all models were similar. The greater

deviations for high and low altitude forest may have arisen from different tree

architectures, wood densities and range of tree diameters used to construct the

equations (Basuki et al., 2009), as well as small sample size (Duncanson et al.,


Page | 163

2015). In contrast, we found very similar means, confidence intervals (95%) and

absolute average deviation when comparing measured AGB with estimated AGB

using our combined forest model and the models of Chave et al. (2005) and

Chojnacky et al. (2014). Similarly, other studies have argued that grouping species

by forest types or eco-regions are an efficient means to estimate AGB, as local AGB

equations will not increase the precision substantially (Chave et al., 2014; Paul et al.,

2016).

Allometric equations developed for specific forest types were more accurate

compared to allometric equations developed for broader eco-regions such as the pan

tropical forest and North American temperate forest. The differences in AGB

predictions from forest specific and eco-regions biomass models are more

pronounced for less diverse forest like the TF and CTF at the lowest and highest

altitude, respectively. However, if specific local or regional biomass equations are

not available, the eco-region models may be sufficient to estimate AGB for larger

areas. We strongly recommend the inclusion of specific gravity of wood in allometric

equations as it led to better model predictions and should be used to construct

allometric equations especially for mixed species forest. The measurement of C

concentration for each species and tree components were necessary for a more

precise estimation of the AGB C stock, as it can vary substantially between species

and within tree components. A holistic estimation of the biomass C needs to include

the estimation of root biomass which was not collected due to time and resource

constraints, and remains a major challenge for biomass C estimation in forest

ecosystems.


Page | 164

References

Akaike, H., 1974. A new look at the statistical model identification. Automatic

Control, IEEE Transactions on 19(6), 716-723.

Arevalo, C.B., Volk, T.A., Bevilacqua, E., Abrahamson, L., 2007. Development and

validation of aboveground biomass estimations for four Salix clones in

central New York. Biomass and Bioenergy 31(1), 1-12.

Baker, T.R., Phillips, O.L., Malhi, Y., Almeida, S., Arroyo, L., Di Fiore, A., Erwin,

T., Killeen, T.J., Laurance, S.G., Laurance, W.F., 2004. Variation in wood

density determines spatial patterns in Amazonian forest biomass. Global

Change Biology 10(5), 545-562.

Banaticla, M.R.N., Sales, R.F., Lasco, R.D., 2005. Biomass equations for tropical

tree plantation species using secondary data from the Philippines, ACIAR

Smallholder Forestry Project-Redevelopment of a Timber Industry Following

Extensive Land Clearing: Proceedings from the End-of-Project Workshop,

pp. 113-128.

Basuki, T.M., van Laake, P.E., Skidmore, A.K., Hussin, Y.A., 2009. Allometric

equations for estimating the above-ground biomass in tropical lowland

Dipterocarp forests. Forest Ecology and Management 257(8), 1684-1694.

Cairns, M.A., Olmsted, I., Granados, J., Argaez, J., 2003. Composition and

aboveground tree biomass of a dry semi-evergreen forest on Mexico’s

Yucatan Peninsula. Forest Ecology and Management 186(1–3), 125-132.




87-99.


Page | 165

Chave, J., Condit, R., Lao, S., Caspersen, J.P., Foster, R.B., Hubbell, S.P., 2003.

Spatial and temporal variation of biomass in a tropical forest: results from a

large census plot in Panama. Journal of Ecology 91(2), 240-252.

Chave, J., Réjou‐Méchain, M., Búrquez, A., Chidumayo, E., Colgan, M.S., Delitti,

W.B., Duque, A., Eid, T., Fearnside, P.M., Goodman, R.C., 2014. Improved

allometric models to estimate the aboveground biomass of tropical trees.

Global Change Biology 20(10), 3177-3190.



Duncanson, L., Rourke, O., Dubayah, R., 2015. Small sample sizes yield biased

allometric equations in temperate forests. Scientific Reports 5, 17153.

Fang, J.-Y., Wang, G.G., Liu, G.-H., Xu, S.-L., 1998. Forest biomass of China: An

estimate based on the biomass-volume relationship. Ecological Applications

8(4), 1084-1091.



49.

Good, N., Paterson, M., Brack, C., Mengersen, K., 2001. Estimating tree component

biomass using variable probability sampling methods. JABES 6(2), 258-267.

Goodman, R.C., Phillips, O.L., Baker, T.R., 2014. The importance of crown

dimensions to improve tropical tree biomass estimates. Ecological

Applications 24(4), 680-698.

Gregoire, T.G., Valentine, H.T., Furnival, G.M., 1995. Sampling methods to estimate

foliage and other characteristics of individual trees. Ecology 76(4), 1181-

1194.


Page | 166

Houghton, R.A., Skole, D.L., Nobre, C.A., Hackler, J.L., Lawrence, K.T.,

Chomentowski, W.H., 2000. Annual fluxes of carbon from deforestation and

regrowth in the Brazilian Amazon. Nature 403(6767), 301-304.

Jessen, R.J., 1955. Determining the fruit count on a tree by randomized branch

sampling. Biometrics 11(1), 99-109.

Levy, P.E., Hale, S.E., Nicoll, B.C., 2004. Biomass expansion factors and root :

shoot ratios for coniferous tree species in Great Britain. Forestry 77(5), 421-

430.

Martin, A.R., Thomas, S.C., 2011. A reassessment of carbon content in tropical trees.

PloS one 6(8), e23533.

Mazerolle, M.J., 2004. Appendix 1: Making sense out of Akaike’s Information

Criterion (AIC): Its use and interpretation in model selection and inference

from ecological data. Mouvements et reproduction des amphibiens en

tourbières perturbées, 174-190.

Monserud, R.A., Huang, S., Yang, Y., 2006. Biomass and biomass change in

lodgepole pine stands in Alberta. Tree Physiology 26(6), 819-831.

Moundounga Mavouroulou, Q., Ngomanda, A., Engone Obiang, N.L., Lebamba, J.,

Gomat, H., Mankou, G.S., Loumeto, J., Midoko Iponga, D., Kossi Ditsouga,

F., Zinga Koumba, R., Botsika Bobé, K.H., Lépengué, N., Mbatchi, B.,

Picard, N., 2014. How to improve allometric equations to estimate forest

biomass stocks? Some hints from a central African forest. Canadian Journal

of Forest Research 44(7), 685-691.



157-166.


Page | 167

Negi, K.S., Rawat, Y.S., Singh, J.S., 1983. Estimation of biomass and nutrient

storage in a Himalayan moist temperate forest. Canadian Journal of Forest

Research 13(6), 1185-1196.

Nogueira Júnior, L.R., Engel, V.L., Parrotta, J.A., Melo, A.C.G.d., Ré, D.S., 2014.

Allometric equations for estimating tree biomass in restored mixed-species

Atlantic Forest stands. Biota Neotropica 14, 1-9.



Paul, K.I., Roxburgh, S.H., Chave, J., England, J.R., Zerihun, A., Specht, A., Lewis,

T., Bennett, L.T., Baker, T.G., Adams, M.A., Huxtable, D., Montagu, K.D.,

Falster, D.S., Feller, M., Sochacki, S., Ritson, P., Bastin, G., Bartle, J., Wildy,

D., Hobbs, T., Larmour, J., Waterworth, R., Stewart, H.T.L., Jonson, J.,

Forrester, D.I., Applegate, G., Mendham, D., Bradford, M., O'Grady, A.,

Green, D., Sudmeyer, R., Rance, S.J., Turner, J., Barton, C., Wenk, E.H.,

Grove, T., Attiwill, P.M., Pinkard, E., Butler, D., Brooksbank, K., Spencer,

B., Snowdon, P., O'Brien, N., Battaglia, M., Cameron, D.M., Hamilton, S.,

McAuthur, G., Sinclair, J., 2016. Testing the generality of above-ground

biomass allometry across plant functional types at the continent scale. Global

Change Biology 22(6), 2106-2124.





Picard, N., Saint-Andre, L., Henry, M., 2012. Manual for building tree volume and

biomass allometric equations: from field measurements to prediction. Food

and Agricultural Organization of the United Nations, Rome and Centre de

Cooperation International en Recherche Agronomique pour le Development.

Montpellier.


Page | 168

Rana, B.S., Singh, S.P., Singh, R.P., 1989. Biomass and net primary productivity in

Central Himalayan forests along an altitudinal gradient. Forest Ecology and

Management 27(3–4), 199-218.

Saatchi, S.S., Harris, N.L., Brown, S., Lefsky, M., Mitchard, E.T.A., Salas, W.,

Zutta, B.R., Buermann, W., Lewis, S.L., Hagen, S., Petrova, S., White, L.,

Silman, M., Morel, A., 2011. Benchmark map of forest carbon stocks in

tropical regions across three continents. Proceedings of the National

Academy of Sciences 108(24), 9899-9904.

Singh, V., Tewari, A., Kushwaha, S.P.S., Dadhwal, V.K., 2011. Formulating

allometric equations for estimating biomass and carbon stock in small

diameter trees. Forest Ecology and Management 261(11), 1945-1949.

Skovsgaard, J.P., Vanclay, J.K., 2008. Forest site productivity: a review of the

evolution of dendrometric concepts for even-aged stands. Forestry 81(1), 13-

31.




Sprugel, D., 1983. Correcting for bias in log-transformed allometric equations.

Ecology 64(1), 209-210.

Symonds, M.R.E., Moussalli, A., 2011. A brief guide to model selection, multimodel

inference and model averaging in behavioural ecology using Akaike’s

information criterion. Behavioral Ecology and Sociobiology 65(1), 13-21.


forests along an altitudinal gradient in the eastern Himalayas and a meta-

analysis of global data. Global Change Biology 22(6), 2255-2268.

Teobaldelli, M., Somogyi, Z., Migliavacca, M., Usoltsev, V.A., 2009. Generalized

functions of biomass expansion factors for conifers and broadleaved by stand


Page | 169

age, growing stock and site index. Forest Ecology and Management 257(3),

1004-1013.




Forests 3(2), 332-352.

UNFCCC, 2007. Report of the conference of the parties on its thirteenth session,

held in Bali from 3 to 15 December 2007, United Nations Framework

Convention on Climate Change (UN FCCC).

Valentine, H.T., Hilton, S.J., 1977. Sampling oak foliage by the randomized-branch

method. Canadian Journal of Forest Research 7(2), 295-298.

Valentine, H.T., Tritton, L.M., Furnival, G.M., 1984. Subsampling trees for biomass,

volume, or mineral content. Forest Science 30(3), 673-681.



broadleaved forest. Journal of Renewable Natural Resources of Bhutan 5(1),

55-67.

Zanne, A.E., Lopez-Gonzalez, G., Coomes, D.A., Ilic, J., Jansen, S., Lewis, S.L.,

Miller, R.B., Swenson, N.G., Wiemann, M.C., Chave, J., 2009. Data from:

Towards a worldwide wood economics spectrum. Dryad Data Repository.


Page | 170

Supplement 6.1 Scatter plots for measured log tree biomass versus log DBH for

a) TF, b) STF, c) WTBLF, d) CTBLF and e) CTF

Supplement 6.2 Tree species specific gravity of wood and C content in wood and

leaves.

Species n S G Wood C (%)

Sun leaf C (%)

Shade leaf C (%)

Tetrameles nudiflora -3 0.35 47.33 48.1 47.84

Kydia calycina -1 0.37 44.59 44.28 44.56

Dubanga grandiflora -2 0.42 46.56 48.97 48.34 Albizzia -3 0.52 46.02 48.8 47.89 Talauma hodgsonii -3 0.59 45.85 48.14 47.43 Pterospermum acerifolium -3 0.59 46.52 50.44 48.93

Bauhinia purpurea -2 0.66 43.91 47.49 45.8 Syzygium -5 0.77 45.43 50.65 50.7 Engelhardtia spicata -1 0.46 46.73 51.08 50.57 Callicarpa arborea -2 0.52 47.14 51.41 50.77 Terminilia -1 0.56 45.66 46.77 45.5

Schima wallichii -6 0.68 47.22 52.02 51.47

Myrica -1 0.85 46.03 - 49.76

Michealia doltsopa -1 0.34 46.76 50.1 48.96

Macaranga indica -1 0.4 39.59 50.15 49.72 Alcimandra carthcartii -1 0.44 46.47 49.56 - Nyssa javanica -3 0.46 45.68 49.84 50.11 Litsea sp. -7 0.48 46.79 51.75 51.31 Eleocarpus -2 0.49 46.29 49.26 48.71 Beilschmiedia -1 0.57 46 51.28 48.93

Viburnum erubescens -2 0.57 46.31 50.55 48.58 Cinnamomum -2 0.64 45.27 51.02 50.22


Page | 171

Castanopsis indica -4 0.7 46.62 49.43 49.38 Quercus lamellosa -1 0.83 44.88 51.45 51.49 Prunus -3 0.46 46.77 50.27 48.79 Schefflera impressa -1 0.48 46.96 56.92 51.38 Magnolia campbellii -2 0.49 46.44 47.56 47.33 Acer sp. -4 0.5 46.42 49.27 47.85 Symplocus sp. -6 0.51 46.06 42.99 44.5 Daphniphyllum charteceum -2 0.55 46.01 49.26 49.41

Rhododendron hodgsonii -7 0.58 46.68 51.04 50.73

Persea sp. -18 0.61 46.5 50.88 50.46 Eurya -4 0.62 47.85 48.05 47.49 Ilex dipyrena -2 0.65 45.36 50.93 49.82 Lithocarpus pachyphyllus -6 0.79 45.69 50.95 51.67

Abies densa -17 0.45 46.44 51.87 52.13 Juniper recuva -5 0.55 47.47 52.6 52.13 Rhododendron arboreum -3 0.55 47.5 51.92 52.42

Numbers in brackets are the number of trees that were sampled. All samples were collected

at 1.3 m above ground with an increment borer (5.15 mm diameter). Leaves were collected

from the sun exposed and shaded portion of the tree after it was felled.

Supplement 6.3 Discrepancies in aboveground biomass C stock estimation for trees

harvested with the use of measured C concentration and considered as 50% of total

biomass.

Forest n

Aboveground biomass C content (kg)

Mean ± SD C% determined

C 50% of biomass

% difference

TF 26 414.0 ± 411.7 449.8 ± 447.3 −8.65 STF 28 709.7 ± 590.2 758.1 ± 630.5 −6.82 WTBLF 31 488.9 ± 456.7 530.5 ± 495.6 −8.51 CTBLF 30 505.4 ± 481.1 545.5 ± 519.3 −7.93 CTF 25 628.0 ± 550.3 671.3 ± 588.2 −6.89 Total 140 547.5 ± 504.7 589.4 ± 542.2 −7.65


Page | 172

Supplement 6.4 Biomass weight of individual tree components of the different

forest types.

sl.no Species Forest DBH (cm) Ht (m) Bole (kg) Brach (kg)

Foliage (kg)

1 Dubanga

grandiflora TF 53 22 896 260 153

2 Sterculiacia sp. TF 39 14 592 161 19

3 Pterospermom

acerifolia TF 42 22 866 38 19

4 Bauhinia purpurae TF 29 20 384 25 17

5 Pterocarpus sp. TF 15 12 35 31 13

6 Lithocarpus sp. TF 18 15 131 48 30

7 Schima wallichii TF 24 10 101 19 16

8 Albizzia sp. TF 44 16 317 105 68

9 Schima wallichii TF 25 13 134 26



12 Albizzia sp. TF 39 21 412 116 31

13 Terminalia

tomentose TF 27 18 304 114 45

14 Tetrameles nudiflora TF 52 25 994 61 55


16 Talauma hodgsonii TF 31 15 307 59 98

17 Bauhinia purpurae TF 18 11 42 3 4

18 Albizzia sp. TF 10 10 23 2 6


20 Sygyzium sp. TF 70 23 1849 1156 178

21 Sygyzium sp. TF 49 20 1585 217 66

22 Sygyzium sp. TF 48 22 1467 380 62



Page | 173


25 Kydia sp. TF 34 18 250 36 169

26 Sygyzium sp. TF 30 18 423 103 176

27 Terminilia sp. STF 62 23 1884 314 144

28 Persea sp. STF 44 25 946 671 678

29 Nyssa javanica STF 75 33 2259 446 89

30 Persea sp. STF 60 29 2236 1705 1059

31 Castanopsis indica STF 18 8 42 8 28

32 Syzyium sp. STF 35 17 841 172 285

33 Myrica sp. STF 19 19 158 47 82

34 Lithocarpus elegans STF 77 27 1591 608 491

35 Persea sp. STF 18 8 79 123 45

36 Persea sp. STF 58 28 2006 258 44


38 Eurya accuminata STF 16 7 47 43 22

39 Eurya sp. STF 29 22 403 22 26

40 Persea sp. STF 20 13 208 38 43

41 Persea sp. STF 34 17 468 132 129

42 Litsea sp. STF 18 15 121 28 84

43 Schima wallichii STF 52 32 1298 224 158

44 Luculia sp. STF 30 14 213 72 88

45 Engelherdia spicata STF 52 24 545 725 356

46 Callicarpa arborea STF 34 18 337 77 58

47 Callicarpa arborea STF 48 15 437 273 247

48 Persea sp. STF 46 23 697 349 94

49 Dubanga

grandiflora STF 51 27 1109 95 116

50 Persea sp. STF 57 29 1491 654 337


Page | 174

51 Persea sp. STF 35 19 508 259 160

52 Eurya sp. STF 35 17 416 192 418


54 Nyssa javanica STF 56 22 928 514 639

55 Virbunum

erubescens WTBLF 11 6 13 2 21

56 Persea sp. WTBLF 41 15 622 152 61

57 Symplocus sp. WTBLF 21 11 70 62 153

58 Virbunum

erubescens WTBLF 32 17 252 108 147


60 Beilschmedia sp. WTBLF 53 28 1093 625 755

61 Litsea sp. WTBLF 20 9 78 21 20

62 Eleocarpus sp. WTBLF 62 26 1499 423 147

63 Alcimandra

carthcartii WTBLF 65 26 1855 107 247

64 Persea sp. WTBLF 65 23 2117 210 260

65 Alnus nepalensis WTBLF 42 22 663 111 252

66 Macaranga indica WTBLF 45 22 538 225 57

67 Litsea sp. WTBLF 24 15 136 23 33


69 Persea sp. WTBLF 18 15 88 5 38

70 Cinnamomum sp. WTBLF 36 28 859 69 54

71 Persea sp. WTBLF 46 18 720 116 116

72 Litsea sp. WTBLF 63 20 1415 446 536

73 Cinnamomum sp. WTBLF 48 23 1221 133 189

74 Litsea sp. WTBLF 36 15 424 85 50

75 Eleocarpus sp. WTBLF 65 23 1600 242 96

76 Unknown (Barkulae) WTBLF 26 15 162 47 30


Page | 175

77 Litsea sp. WTBLF 19 13 82 7 17

78 Nyssa javanica WTBLF 30 13 241 47 29

79 Daphniphyllum

charteceum WTBLF 24 15 139 55 124

80 Litsea sp. WTBLF 53 22 1230 162 73

81 Acer sp. WTBLF 35 17 471 26 20

82 Michealia WTBLF 44 20 721 219 65

83 Quercus lamellosa WTBLF 56 22 1829 407 156

84 Persea sp. WTBLF 47 19 608 96 39

85 Castanopsis indica WTBLF 68 22 2669 791 479

86 Daphniphyllum

charteceum CTBLF 36 14 420 73 30

87 Eurya sp. CTBLF 33 15 303 109 52

88 Acer sp. CTBLF 54 22 1576 92 118

89 Symplocus sp. CTBLF 22 15 65 39 24


91 Persea sp. CTBLF 44 19 920 138 108

92 Prunus sp. CTBLF 33 19 185 92 52

93 Lithocarpus sp. CTBLF 64 23 1777 990 127

94 Ilex dipyrena CTBLF 44 18 998 110 109

95 Magnolia campbellii CTBLF 35 17 398 261 97

96 Rhododendron sp. CTBLF 56 17 1212 446 715

97 Lithocarpus sp. CTBLF 58 26 1723 130 91

98 Prunus sp. CTBLF 19 13 64 15 8



101 Schefellera CTBLF 35 17 296 35 34

102 Lithocarpus CTBLF 67 25 2798 531 573

103 Persea sp. CTBLF 37 16 325 193 156


Page | 176

104 Magnolia campbellii CTBLF 53 20 1057 427 245

105 Acer sp. CTBLF 63 24 2074 457 357



108 Persea sp. CTBLF 23 14 132 18 10

109 Ilex dipyrena CTBLF 55 19 1798 158 230

110 Persea sp. CTBLF 68 20 2084 258 101



113 Prunus sp. CTBLF 24 14 164 87 110



116 Rhododendron

arboreum CTF 18 10 77 11 18

117 Rhododendron

arboreum CTF 14 9 27 49 33

118 Rhododendron

arboreum CTF 26 12 118 58 50

119 Juniper recuva CTF 24 15 211 21 14


121 Abies densa CTF 47 16 1556 148 77

122 Abies densa CTF 39 18 1009 140 75

123 Abies densa CTF 58 18 2503 800 527

124 Abies densa CTF 64 20 2056 564 655

125 Abies densa CTF 66 21 1579 504 633

126 Abies densa CTF 63 18 2308 323 232

127 Abies densa CTF 35 14 557 62 29




Page | 177


131 Abies densa CTF 60 16 2062 526 269

132 Abies densa CTF 55 17 2035 447 228

133 Abies densa CTF 30 14 393 62 70

134 Abies densa CTF 42 19 927 226 255

135 Abies densa CTF 38 15 670 31 49

136 Abies densa CTF 35 12 421 48 56

137 Abies densa CTF 34 14 820 87 57

138 Abies densa CTF 20 13 134 8 11

139 Abies densa CTF 25 14 241 25 36

140 Abies densa CTF 35 14 549 123 26

Chapter 7. General Discussion and Conclusions

Page | 178

Chapter 7.

General Discussion and Conclusions

This thesis presents comprehensive data for forest carbon (C) and nitrogen

(N) stocks in soil and vegetation along an altitudinal gradient in the eastern

Himalayan regions in Bhutan (Chapter 3). The transect used for this research

encompassed five different forest zones, namely tropical forest (TF), sub-tropical

forest (STF), warm temperate broadleaf forest (WTBLF), cool temperate broadleaf

forest (CTBLF) and cool temperate forest (CTF). Soil C and N stocks for the top 100

cm, increased from 114 Mg ha-1 at the lowest altitude (TF), to 403 Mg ha-1 at the

highest altitude (CTF). The increased soil organic carbon (SOC) content with

increasing altitude is in agreement with the meta-analysis of published data

performed as part of this thesis. Along the study transect soil C and N stocks were

positively correlated with CEC, species richness of the forest and quantity of litter

input in the soils. In contrast, soil C and N concentrations were negatively correlated

with bulk density and pH of soil. The C concentrations in understorey biomass also

increased with altitude, e.g. foliage %C rose from 39.7 to 45.7% with a rate of 2%

for 1000 m rise in altitude. Rising C concentration in biomass might have been an

additional factor for increased SOC with altitude. Deep soils (between 60 − 100 cm)

stored 24 to 36% C and 22 to 25% N stocks of the total C and N considered for the

whole profile of 100 cm. This constitutes a substantial amounts of C and N and needs

to be considered for an accurate C stock estimation, as many studies only consider

depths to 60 cm. Along the soil depth profile, deeper soils at lower altitudes had

lower C:N ratios, suggesting advanced decomposition and older SOM in deeper

soils. Deeper soils are more important for C accounting as deep soil C has a

relatively longer residence time due to nutrient and energy limitation in the sub-soils

(Fontaine et al., 2007). With increasing altitude, not only C and N stocks, but also

C:N ratio of soils increased. This suggests that the large C stock in high altitude

forest is mainly due to reduced decomposition with increasing altitude and


Page | 179

decreasing temperature. C and N stocks in high altitude forest soils were amongst the

highest reported in the world.

To understand input processes, turnover of OC along the studied transect, C

and N stable isotopes and elemental contents of biomass and soil were used (Chapter

4). Soil δ15N was stable along the altitudinal gradient, whereas δ15N in all categories

of biomass decreased with increasing altitude with a gradient of 0.5‰ to 0.8‰ per

km. This could be due to a ‘less open’ N cycle at higher altitude forest sites (Karolien

et al., 2013). The decreasing trend of δ15N was not observed in soil, which suggests

that total soil N was not representative of the N taken up by plants. The difference

between plant and soil δ15N (Δδ 15Nplant-soil) became larger with altitude, which may

have been the result of a shift of N-source from NO3− in tropical forests to organic

NH4+ N in the cooler high altitude forests (Amundson et al., 2003).

δ13C for the understorey vegetation were comparatively more depleted than

the overstorey trees (−31.7‰ to −29.9‰) due to light limitation (Gessler et al.,

2004). The δ13C values for the overstorey biomass components depicted a curvilinear

trend with altitude and δ13C was greatest in mid-altitude forests. The initial

decreasing trend of δ13C in the overstorey biomass was attributed to decreasing light

intensity with increasing altitude up to altitudes of c. 1800 m due to persistence and

intensity of fog in the study area. As light conditions were not measured directly,

relative humidity was taken as a proxy for fog (Syed et al., 2012), and the mean

annual relative humidity for the study site (Chapter 4, Table 4.1) corresponded with

the observed pattern of fog. The pattern of δ13C at all soil depths corresponded with

the pattern of δ13C in the overstorey biomass, which indicates that overstorey

biomass was the major contributor to soil C along the transect.

Along the soil depth profile, the rate of δ13C and δ15N enrichment was least

for the highest altitude CTF. Further, high altitude CTF had the least regression slope

of soil δ13C to logarithm of C concentration, which is indicative of abundance of 13C

related to the degree of organic matter decomposition (Garten, 2006). These confirm

that organic C turnover is slowest in the soils at the highest altitude forest.


Page | 180

Apart from the quantity of C stored in the soil, it is also important to

determine the proportion of C forms and their association with clay minerals

(Chapter 5) to determine the longevity of C. Carbon forms, organo–mineral

associations and environmental conditions affect SOC storage and stability.

Particulate organic matter proportion was greater in the lower density fractions and at

greater altitude forests. The relatively high C:N ratio in the POM suggests high

content of relatively fresh biomass which is at an early stage of decomposition

(Golchin et al., 1994). In contrast, heavier density fractions (> 1.8 g cm-3) were

present in greater proportion in lower altitude soils. With increasing density, the δ13C

became enriched and the degree of enrichment was smaller at higher altitudes. These

results indicate that more microbially processed SOC was stored in the heavier

density fractions (Baisden et al., 2002; Jones and Singh, 2014; Nadelhoffer and Fry,

1988), and the decomposition rates decreased at higher altitudes.

Particulate organic matter and lighter density fractions had a greater

proportion of phyllosilicates which were associated with increased content of

aliphatic C. In contrast, increased aromatic C was associated with increasing density

fractions. The increased proportion of aromatic C as compared to aliphatic C with

increasing density indicates that C in the heavier density fractions were more

decomposed (Baisden et al., 2002). The proportion of aliphatic C increased and

aromatic C decreased with increase in altitude. This corresponded with significantly

smaller Index 1, a metric of decomposition, for higher altitude forests.

Correspondingly Index 2, the relative recalcitrance for SOC, at the highest altitude

(CTF) is low, suggesting limited decomposition of even the more easily

decomposable carboxyl group and polysaccharides. Additionally the small change in

δ13C values with increasing density for the CTF supports the hypothesis that

decomposition is limited at the highest altitude forest soils.

In summary, the soil C and N stock increased with altitude of the forest.

However, isotopic analysis revealed that C and N stored in high altitude forest are

mainly due to limited decomposition. From the characterisation of the different

forms of C stored in the different altitude forest soils, the greater proportion of the C


Page | 181

stored in the high altitude forest are the easily decomposable forms of C, the

aliphatic, carboxyl and polysaccharides.

Although SOC forms the major proportion of the terrestrial C stock, forest C

dynamics would not be complete without consideration of the aboveground biomass

(AGB) and C stocks. To estimate the AGB stocks, a comprehensive forest inventory

and measurements of ground vegetation and surface plant residues were made.

Additionally, trees were harvested to develop biomass models for each of the forest

types along the altitudinal gradient (Chapter 6).

To effectively measure the AGB, some of the parameters considered were

wood specific gravity (SG) and C concentration in the different components (e.g

foliage and trunk) of the tree. The average wood SG for the forest types ranged from

0.48 to 0.63 and was the least for the CTF at the highest altitude that were conifer

dominated forests. Wood C concentration for individual trees for the entire study

area ranged from 39.6% to 47.9%, which is below 50% C in biomass, which has

been used in some studies of biomass C stock estimation. Leaf C concentrations for

individual trees ranged from 43.0% to 56.9%. Consideration of the actual C

concentration rather than the assumed 50% C in the biomass to estimate C stocks can

potentially reduce estimation errors in biomass C stocks by 6.8 to 8.6%.

To construct the biomass equations, key variables like tree DBH, height and

SG were considered. AGB and tree parameters were log transformed to develop the

best fit model. Out of the six models with combination of the different key variables,

two (Model 1 and 4) better predicted AGB for the five different forest types and for

combined entire forest along the altitudinal gradient. The selection of the models was

based on Akaike Information Criterion (AIC), root mean square error (RMSE),

coefficient of determination (r2) of the regression and absolute average deviation

from the measured AGB. Models with wood SG as a predictor variable has not been

chosen as the recommended model for the present study mainly because data on SG

is not collected during conventional forest inventory. However, models with SG as a

predictor variable always featured as the best model based on the selection criteria.

This is because SG captures the variability of the measured data especially for


Page | 182

contrasting sites with mixed tree species (Basuki et al., 2009; Chave et al., 2005;

Chave et al., 2014). Thus it is recommended to collect wood samples for SG

measurements during forest inventory, especially if mixed species forests AGB

models are to be used.

Aboveground biomass estimated from the recommended models for specific

forest types provided better estimates compared with other published models (Chave

et al., 2005; Chojnacky et al., 2014). Greater deviation from the models by Chave et

al. (2005) and Chojnacky et al. (2014), may have risen because of the differences in

tree species composition, tree architecture, wood SG, range of diameters used for

model construction (Basuki et al., 2009), and sample size (Duncanson et al., 2015).

However, if specific local or regional AGB equations are not available, the eco-

regions models proposed by Chave et al. (2005) for pan tropical forest and by

Chojnacky et al. (2014) for specific taxa grouping for North America tree species

may be sufficient.

The AGB C stocks for the different forest were TF = 57.5, STF = 91.2,

WTBLF = 112.7, CTBLF = 157.4 and CTF = 207.7 Mg C ha-1. From AGB C stocks

and SOC stocks (Chapter 3) the total forest soil C stocks were estimated. However

there still is a lack of the C stocks estimates for the forest root biomass. Due to

methodological challenges and associated large root recovery errors (Addo-Danso et

al., 2016), relatively little work has been done on the estimation of root biomass. A

synthesis of global upland forests data (independent of latitude, soil texture, or tree

type) show the root shoot ratio to be between 0.20 and 0.30 (Cairns et al., 1997).

However, more efforts need to be directed to estimate roots biomass due to greater

variability in the allocation of C to shoot and root (Cairns et al., 1997).

In conclusion this research on C and N stocks and development of forest

biomass models is the first comprehensive study in the Bhutan Himalayas. Deeper

soils (60 − 100 cm) were found to store a substantial amount of C and N and must be

considered for better estimation of C stocks. Forest specific biomass equations were

better predictors of biomass than general global models. The C stocks in the soils as

well as in the biomass were greatest in the high altitude forests, but C in the soil was


Page | 183

at a stage of limited decomposition and present in easily decomposable forms

(aliphatic, carboxyl and polysaccharides). This may have stark implications in a

hotter climate where high altitude soil C in the Bhutan Himalayas may be easily lost.

These findings are important for any future management of these forests.


Page | 184

References

Addo-Danso, S.D., Prescott, C.E., Smith, A.R., 2016. Methods for estimating root

biomass and production in forest and woodland ecosystem carbon studies: A

review. Forest Ecology and Management 359, 332-351.

Amundson, R., Austin, A.T., Schuur, E.A.G., Yoo, K., Matzek, V., Kendall, C.,

Uebersax, A., Brenner, D., Baisden, W.T., 2003. Global patterns of the

isotopic composition of soil and plant nitrogen. Global Biogeochemical

Cycles 17(1), 1031.



Global Biogeochemical Cycles 16(4).

Basuki, T.M., van Laake, P.E., Skidmore, A.K., Hussin, Y.A., 2009. Allometric

equations for estimating the above-ground biomass in tropical lowland

Dipterocarp forests. Forest Ecology and Management 257(8), 1684-1694.

Cairns, M.A., Brown, S., Helmer, E.H., Baumgardner, G.A., 1997. Root biomass

allocation in the world's upland forests. Oecologia 111(1), 1-11.




87-99.

Chave, J., Réjou‐Méchain, M., Búrquez, A., Chidumayo, E., Colgan, M.S., Delitti,

W.B., Duque, A., Eid, T., Fearnside, P.M., Goodman, R.C., 2014. Improved

allometric models to estimate the aboveground biomass of tropical trees.

Global Change Biology 20(10), 3177-3190.




Page | 185

Duncanson, L., Rourke, O., Dubayah, R., 2015. Small sample sizes yield biased

alometric equations in temperate forests. Scientific Reports 5, 17153.

Fontaine, S., Barot, S., Barre, P., Bdioui, N., Mary, B., Rumpel, C., 2007. Stability of

organic carbon in deep soil layers controlled by fresh carbon supply. Nature

450(7167), 277-280.

Garten, C.T., Jr., 2006. Relationships among forest soil C isotopic composition,

partitioning, and turnover times. Canadian Journal of Forest Research 36(9),

2157-2167.

Gessler, A., Rennenberg, H., Keitel, C., 2004. Stable isotope composition of

organiccompounds transported in the phloem of European Beech - Evaluation

of different methods of phloem sap collection and assessment of gradients in

carbon isotope composition during leaf-to-stem transport. Plant Biology 6(6),

721-729.

Golchin, A., Oades, J., Skjemstad, J., Clarke, P., 1994. Soil structure and carbon

cycling. Soil Research 32(5), 1043-1068.

Jones, E., Singh, B., 2014. Organo-mineral interactions in contrasting soils under

natural vegetation. Frontiers in Environmental Science 2.

Karolien, D., Ilaria Del, G., Andrea, V., M Francesca, C., 2013. Assessment of soil C

and N stocks and fractions across 11 European soils under varying land uses.

Open Journal of Soil Science 2013.



Journal 52(6), 1633-1640.

Syed, F.S., Körnich, H., Tjernström, M., 2012. On the fog variability over south

Asia. Climate Dynamics 39(12), 2993-3005.

copyright and use of this thesis · other degree or diploma of the university or other institute of...

Documents